Yext, Inc. (YEXT) Earnings Call Transcript & Summary

Hi. My name is Yuka Broderick, and I'm Head of Investor Relations at Yext. I'd like to welcome you to the Yext 2021 Investor Day. We have a great lineup of executives to share our vision, strategy and financial performance and targets with you. Let me briefly walk you through our agenda for today. First, CEO, Howard Lerman, will speak to you about how keyword-based search is ripe for disruption. Our Chief Strategy Officer, Marc Ferrentino, will follow with some demos and an explanation of our technology and differentiation. Then we'll have some time for Q&A with Howard and Marc. After a short break, we'll return with President and Chief Revenue Officer, David Rudnitsky, to talk about our go-to-market and share some information about Answers and our platform sales. Finally, CFO, Steve Cakebread, will talk about our growth and productivity drivers and our financial goals. And finally, we'll gather together all of the executives presenting today for another Q&A session. As a reminder, our presentations today will contain forward-looking statements, which do not guarantee future events or performance. These forward-looking statements are subject to certain risks, uncertainties and assumptions, which are discussed in our reports filed with the SEC. We also will be referring to non-GAAP measures. Reconciliations with the most comparable GAAP measures are available in the appendix to these materials, which are posted at investors.yext.com. Also, I'd like to point you to the Q&A submission window in the bottom left-hand corner of your screen. Feel free to type in questions at any time during the presentation. We will select from these submissions in the Q&A sessions. With those instructions out of the way, let's get moving on our Investor Day. Thanks. [Presentation]

Today, I'm going to tell you about the second most important search engine ever built. But first, let's go back to prehistoric times. 1994, a magnifying glass, keyword search, hyperlinks. The year 1994 marked the explosion of keyword search. In a mere 36 months, the Internet saw dozens of keyword search engines, including Infoseek, Yahoo!, Lycos, WebCrawler, LookSmart, Excite and AltaVista. And then in 1998, Google launched PageRank, which turned out to be the best algorithm to rank web results for keyword searches on the consumer Internet. This is where most search stopped improving 23 years ago. Fast forward a year to 1999. Former Xerox PARC Engineer, Doug Cutting, launched an open-source keyword search engine called Apache Lucene. Lucene turned out to be the second most important search engine ever built. With Lucene, enterprise developers could add a keyword search engine to their websites, to their enterprise apps and their help desks, e-commerce sites and more. Now fast forward 5 more years. In 2004, a more developer-friendly open-source keyword search engine built on Lucene, called Solr came along. Solr to date retains huge market share in search. Even as Elasticsearch, its competitor, which is also built on Lucene, gained share. Even today, the odds are that you use Lucene-powered search almost as often as you use Google. You just don't know it. But Google Search is vastly superior to Lucene-powered keyword search, and that's because Google pioneered a breakthrough in search technology called Natural Language Processing or NLP. NLP understands things. It understands questions like when was Marco Polo born? And it tells you. Your searches on Google use this revolutionary new technology. But nearly every other search experience on a website, in an app, in the workplace, on a health site, that's just a simple keyword search based on the same Lucene technology from 1999. These search experiences almost always give you a set of blue link results that rarely helps you find what you want. You know exactly what I'm talking about here. You see the magnifying glass on a website, you type in a keyword, you get a terrible set of irrelevant results back. So you give up and you try something else to help you find what you want. Keyword search is prehistoric. Yet, still everywhere. We think that's crazy. So we invented Yext Answers. Yext Answers is a modern search engine that uses NLP, powered by advanced artificial intelligence to understand questions, uses multiple algorithms to present different sets of results in a dynamic UX and is built on a knowledge graph. Google brought modern search to consumers, Yext is going to bring modern search to the enterprise. Let's compare for a minute, Yext search to outdated keyword search. We're going to compare first the input that the algorithms that compute your answer and then the results. We'll serve the input. Keyword search works like Control F in a word document. You type in your input query, and the keyword search finds all the places in the document that contains the exact string you've asked for. But the problem is that a lot of times, you don't search for keywords that exactly match what you need. Take a look at this, Solr-powered keyword search on a leading HMO company's website. A search for "physicians in Vienna, Virginia" gives you know results back. And that's because there's no place in the entire website that contains the exact string "physicians in Vienna, Virginia." That's a pretty bad experience. Yext search uses AI-powered natural language to fundamentally understand the user's input. We also used named entity recognition to turn an unstructured question into a structured query of the knowledge graph. And this allows for the ability to understand very precise questions. Questions like "physicians in Vienna, Virginia who accept Aetna and speak Spanish." Yext Answers automatically turns this question into a structured graph query to find matching entities right on the customer's knowledge graph. And that's how Google works. And so it enables the user to ask much deeper and much more precise questions. Like Google, Yext search understands phrases. Keyword search would literally take that entire string, "Physicians in Vienna, Virginia, that speak Spanish and accept Aetna," and hit Control F, if you will, to find places in the document that match it. Odds are, there'll be no results, and that is the 1990s approach to search. Okay. Next. Let's compare how keyword search uses a single-ranking algorithm to how Yext Answers uses and blends together multiple algorithms to create multiple result sets. The Control F keyword search approach works really well for a reasonably sized document, like a research paper, even believe it or not, a long 500-page book because the user can just chronologically find next, next, next, next, next, throughout the document. But the world wide web exploded it into billions of pages. And the keyword, Marco Polo, for example, is mentioned on the world wide web in 207 million places. And that is way too many for anyone to sort through. These places need to be ranked. And so keyword search engines of the late '90s adapted by focusing on their ranking algorithm. Now if the keyword, Marco Polo appears in 200 million places, search engines would need to have a strong algorithm that ranked hyperlinks by relevancy. The user could then click-through to the page they felt best matched their keyword. And thus, keyword search evolved to be what we call algorithm first. It's designed around ranking millions of results. Now contrast that to modern search. Modern search uses a list of links only as the last resort. Now take a look at this Google search for the keyword McDonald's. There are 500 million results, but only a single web link at the top, McDonalds.com. So that's the one. The rest are barely visible at the bottom. Google combines snippets and maps and knowledge cards and other elements to present the user with multiple options for their results. And each of those different elements uses a different algorithm. Answers from the knowledge graph, like maps or knowledge cards like this, like we used named entity recognition. Answers in their feature snippets, like this, likely use Extractive QA. That's an algorithm that drives answers from unstructured text. Answers and people also ask this element you see right here, like we use semantic text search to drive answers from semi-structured text. There are many different kinds of elements on every Google search engine results page that get the user what they want. And each element uses a different algorithm. This is how Yext Answers works. For each query that comes in, we apply multiple algorithms to the question and give the user back a dynamic result that best matches what they've asked for. Keyword search uses one algorithm to rank all the millions of results. Yext Answers, like Google, combines multiple elements in a dynamic UX to give the user the best experience. Web results, they're just to fall back when we can't find anything else. We use them as the last resort. Finally, let's compare results. Keyword search literally provides a list of hyperlinks ranked by relevance. You have to click the link and then you have to read the results. This does not complete your quest. It just points you in the right direction. Yext search completes the search loop by answering your question exactly, and it allows you to transact as the next step. Users get multiple elements to answer their questions. Sometimes, it's a list from the knowledge graph, like match or listed people. Sometimes, it's direct answers from a knowledge graph. Sometimes, it's snippets from Extractive QA. And best of all, when the user gets their results, they can transact right off the search engine results page. Depending on what you ask, your result set offers different kinds of transactions. You can order online. You can request directions or make a call or buy a product, keyword search gives you hyperlinks, Yext search gives you dynamic answers. So ladies and gentlemen, there you have it. Yext search is better than keyword search on literally every dimension. With keyword search, your quest continues. With Yext Answers, your quest is over. But we're not just better. It turns out we're also faster and also cheaper. Let's talk about this for a second. In order to implement a Lucene-powered keyword search engine from the '90s, it requires a lot of expertise from specialized developers who know quite a lot about search. There are definitely some companies that ought to build their own custom search engine for their core experience. Spotify or Home Depot or Slack come to mind. Almost every company needs some kind of search. But most companies in most places shouldn't build a search engine. A better approach is just to turn on Yext search-as-a-service, much like you might turn on Stripe for billing or Twilio for communications, admins and developers still have control versus a black box, like Google, but it's low code and a developer can turn it on in basically a day. And this means costs are low and time to value is fast as it's now. Better, faster, cheaper. Yext Answers dominates keyword search on every dimension. Our mission as a company is perfect answers everywhere. And everywhere in the enterprise means roughly 5 categories in our search platform. Website search, where we already are with over 240 customers and 25 million searches in Q4, support search. That's next. This is search you might see on a health site. And our multiple-algorithm approach, especially with Extractive QA can help companies answer their customer questions and deflect support calls. We intend to enter this market in H1 of this year. So about app search. That's the third category. This lets developers build Yext search directly into their products. Startups and tech companies can easily get a modern natural language search going, really easily. And we'll be competing in this market in H1 of this year. E-commerce search lets companies sell products in their sites. And we are already seeing a trickle of demand here. We'll be in this category later this year. Workplace search. This is a big one, behind the firewall. It's actually the #1 request we received from CIOs and CTOs. We hear, "Your site search is spectacular. But when can we use this for enterprise search needs?" Well, look to us to be ready to help serve that need later this year. As we land with one search experience in a line of business, we'll be able to expand with other search solutions to our customers. You heard me talk a lot about our vision to disrupt keyword search with better, faster, cheaper answers. Now it's important to keep in mind that we are well positioned to bring natural language search to the enterprise because of our dominant and growing listings franchise. Let's take a step back. Yext was founded in 2006 with the mission of putting perfect answers everywhere. Our founding principle was that, and still remains, the ultimate authority of information about a business is the business itself. This just makes sense that Tesco in the United Kingdom is the authority on where their stores are or Wendy's is the authority on nutritional information about a Frosty. This basic principle led our -- led to our breakthrough listings innovation. It's kind of a simple idea. Let companies control all their own information and services, like Google and Apple and Amazon. And these companies, these publishers gave Yext special access to update business information, like addresses and menus, but we had to give them all this data, millions of facts in a structured way. We couldn't just send them a flat file of text with all these facts. We had to define what entities exist and what fields are on those entities and what values are populating those fields, and most importantly, how those fields and entities matched up to the data in these external services. This is actually pretty hard. It's really complicated. So our patented match-and-lock listings technology was born to link up the tens of millions of facts in Yext into all these services directly. So we built a knowledge graph to mirror our data to the knowledge graphs in other services like Google. Every time a Yext customer uses listings, they add the data to their own knowledge graph, and we send it to the right cell everywhere else. One day, a few years ago, we thought, we're setting all these answers to Google and to Apple. Why not let our customers answer questions themselves? And so Answers was born right out of listings. Listings built on our knowledge graph platform is a core growth franchise for Yext. And obviously, it faced extreme macroeconomic challenges here. In Q4, Google Maps listings were down more than 50% on a per location basis versus the prior year due to COVID-19 related lockdown -- location shutdowns. But despite this, the product still grew in 2020, and that is a testament to its criticality and its resiliency. I don't think anyone thinks that Google Maps volume will be permanently less than half of what it was in the previous year. So we have some really cool stuff coming out for listings. As part of our expansion into the e-commerce category, we'll be introducing product needs. We continue to innovate on reviews. Our partners are hungry for new types structured data objects and fields, and we'll be supporting them. And plus their new search engines and services gaining traction that are going to need our structured data, showing up with 475 million structured facts that are from the primary source is pretty compelling for a new search engine, and that gives us a huge head start in working with them. And in the past year, we added 28 new listings partners to our network, including DoorDash, WebMD and OpenTable. We also upgraded tons of existing integrations. Our customers can now share types of structured data, like curbside pickup options, delivery, drive-through hours, temporarily closed flags, online events in services like Apple Maps and Google My Business and Bing and Nextdoor and Facebook. Search is evolving. DuckDuckGo is exploding. Snap Maps is incredibly promising. Evan just talked about that. As long as people use search engines, maps, intelligent agents, they need car directions, there's going to be a need for a clearing house of primary structure data. And this gives Yext an incredible position as we gear up to get back to business. With listings and answers together, Yext has a dominant offering in intelligent search. Enterprises can build a single-structured knowledge graph for their company and AI-powered answers appear everywhere in smart services, like Google and Siri, and within natural language search and their own websites, their own support sites, e-com, everywhere. Everywhere is indeed the right word to describe search because search is in nearly every app, every site, your phone, your car, that magnifying glass is everywhere. 15 years ago, when we founded Yext in a cold single room on the Upper West side, search was all about keywords, ranking blue hyperlinks. But the search of tomorrow is intelligent. Magnifying glasses are now voice input boxes. Keywords turned to questions, bright blue links are now answers. The massive platform shift from keyword search to natural language search is underway. Our opportunity ahead has never been greater. We estimate our total addressable market to be at about $30 billion by 2024. And our objective is to be the worldwide leader in structured data and AI-powered natural language search for the enterprise. For more than a decade, Yext has innovated to put perfect answers everywhere. We always have and we always will. And now I am elated to introduce our Chief Strategy Officer, Marc Ferrentino, to give you a demo of our latest and greatest technology.

Thanks, Howard. I'm Marc Ferrentino, Chief Strategy Officer here at Yext. And I would like to talk with you about innovation at Yext. We are accelerating innovation. We are far outpacing our competitors right now in the market. I want to specifically talk about our Spring '21 Release that just came out this morning. It's full with almost 65 new features. Some of the highlights are document search. This is a brand-new algorithm that we're adding to our multi-algorithm lineup that allows you to go into an unstructured document and pull the answer directly out of it. The answer could be a word, it could be a sentence or could even be a paragraph. I want to talk about our data connectors and crawler. This is a brand-new feature. One of the biggest things we've heard from our customer base is, "I love the knowledge graph, but how do I build one?" And so our data connector framework and crawler is built and made so that people can easily build knowledge graphs on the fly. I want to talk about the launch of our SDK and our Command Line Interface, a brand-new developer persona that we're putting out through our Hitchhiker program. And of course, authenticated users. One of the biggest requests that we've gotten is, "I want answers behind the firewall. I want to use it inside my company, not only outside my company." And so what we've done is we've upgraded our security at a ton of new features, like an encryption at rest, like token-based API calls. And now we actually can handle all these different use cases even behind the firewall. So I want to take a minute here and actually do a demo of some of this technology that we put out over the last 6 months. So today, I'm going to show you how to build the search experience for Sabah. Sabah is a New York-based shoe company that prides itself on its handcrafted shoes for men and women that use very high-end materials. As you can see, they don't currently have search on their website. So we're going to build the search experience for them. Now if you look at the content on the website, what you'll see is you'll see products, you'll see stores, you have FAQs and they even have something called a journal, which is like a blog post. So we're going to make all this searchable, natural language on the Answers platform. So the first thing we need to do is get this content into Yext. We store content in something called a knowledge graph. A knowledge graph is a database that maintains the semantic structure of the data, allowing us to deliver better search. So let's jump into the Yext platform and show you what I mean. Now ahead of this demo, we started to build out some of the graph. We added entities for products, for locations, for FAQs, and we actually ended a few articles that were on a different section of the website. Now you can add more entity types if you want to continue making this experience better. And there are a lot of different ways to get data into knowledge graph. You can click on this add button here and add data in by entering a single entity, uploading a CSP, connecting with other systems like Shopify and Magento. By the way, we'll be adding ServiceNow, Zendesk and Service Cloud as part of this release or you can use our new crawler. That's part of our spring release. So you can see that I've already populated all of the products. There's 1,251 of them. I did this using the Shopify Connector. I put in all 3 locations by hand. And I used our crawler to bring in all the FAQs crawling the FAQ page that was on the Sabah website. Now the crawler makes it very easy to pull in content into the knowledge graph. So let me show you how it works. So we've already loaded 4 articles from -- up from another connector, but there is actually some additional content on the Sabah site under Journal, as I mentioned before. In here, you'll see about 5 journal entries. So I'm going to show you how you would add them into the graph. Well, the first thing you would do is you would click on add data. And in here, you would create a connector. A connector is a combination of 2 things. There is the source and then of course, there is the actual object that we're bringing the data into. And you can see here, we've previously crawled the entire website, so we've already got the pages loaded in the system. And now we're going to create a connector that extracts just those journal pages. So since we just load the journal pages, we're going to go in, and we're going to set up a filter that takes -- it takes just the 5 journals, and we're going to use a star character here as a way of illustrating that. We hit save and continue. Now what we show you here is a preview of some of the information that we've already parsed in. So things like page ID, the URL, page title. And you can see it's got the page titled the journal, it's got the URL, the actual journal link itself. And what we're also going to want here is we're going to want to add the body of the journals themselves. So let's take a whole new field called page body. And we're going to pull in all of the clean text from the HTML page itself. And you can see here there it is. This is literally all the text cleaned, ready to go, pulled out of the body. So now we have everything we want from the source. And now we're going to move on to the next step of mapping. So our NVID is already mapped to the page ID. And so next -- what we're going to do next is we're going to map the website URL to the URL object or the URL field. We're going to map the page title to the name field, and we're going to map the body to the body field. So now we have our mappings, we can now save our connector. And you see here is our connector. You can rename it from site connector to article connector. And then we will run it now. So what we're doing is now processing the information, pulling and extracting the information out of each page, and you see very quickly, we created 5 brand-new entities. So if you go back to the knowledge graph, you'll see that the article number went from 4 to now 9 entities. And you can see these journals are now inside the knowledge graph here with the full body of text. So as you can see, building a knowledge graph is pretty easy, and you can easily add new content as your content changes. I'd like to show you how to set up an Answers experience and how the multi-algorithm approach really improves search results of classic keyword search to give your customers that modern Google-like search experience. So I set up a simple experience ahead of time for each of the entities in the knowledge graph, like product, location, FAQ and articles. Here's what the user experience would look like. And as you can see, the tabs and the screens are mapped up to each entities. Now let's jump over to the Yext admin screen. And you can see the algorithm configuration for each of the entities that we're talking about. Now we set each one of these entities up to keyword search, just to start off. So let's take a first spin and see what it can do. So let's start by searching for NYC. Okay. So that's good. But the reason it's coming up is that New York City is in the name of the location. So that's fine if someone is just looking for New York City, but what if they're searching for New York? Didn't get anything. What if they're searching for stores near me? Not much here. What about something really complicated like phone number of the New York location. So as you can see, not a great experience. This is the downside of keyword search. It's not smart enough to understand the query. So let's go back and turn on one of our natural language algorithms. We're going to turn on the NLP filter algorithm, which is our Named Entity Recognition algorithm just for this location entity. We are now going to be able to do geo searches. And that was all you had to do. You just had to literally change to one of the NLP AI algorithms and then click save. Now let's go back and let's try some of those same searches. So New York City? Okay. Great. That works. New York? Well, that works, too. Stores near me? Okay. We get this big map here. And actually, since I'm based in Miami, it actually is zooming out to the entire -- to show the 3 stores across the United States. And let's try that a very complicated one, phone numbers of New York location. Oh, wow, look at that, it gave an exact answer. The exact phone number of the New York location. In this instance, we not only found the most relevant store, but we also found the most relevant field inside of the location object. FAQs are very common. Right now, our FAQ entity has keyword search turned on. So let's try some searches. Shall we? Let's try waterproof. Okay. So that's good. We get some results back. But very much like the location search, it's actually matching against the keyword waterproof. Let's try something a little more interesting. Let's try water resistant. Okay. Not much there. Okay, let's try something a little further, a little crazy. Can I step in the puddle? That's a pretty good one. And once again, no results. Okay. Let's go back and turn on the semantic text algorithm for the FAQ entity. Once again, I turn off keyword search. I turn on semantic tech search. I hit save. So this algorithm works by looking at the semantic embeddings and looking at the space between the query turns and the semantic relevance of the FAQ. So they don't need to have keyboard overlap to show up in the results. Let's take a look at how this impacts the search results. Solr proof. Okay. So we get the same 2 FAQs before, but we actually get another one, which, when you look at it, actually makes sense. Water resistance? Once again, we get 3 very relevant FAQs. And the last one is, can I step in a puddle? And actually, if you look at them, it's actually accurate and pretty dead on. Now semantic text search leverages semantic understanding that is based on a language model that has read the entire Internet and understands that waterproof and water resistant are semantically similar terms. While FAQs are considered semi-structured data, articles are considered unstructured data. If you remember, we crawled the Sabah website for journal entries and put the unstructured text directly in the knowledge graph. We've turned on keyword search for this entity. So let's see how it performs. I'll start by doing a simple search. Resold. I get back an article that has tons of text about all aspects of reselling your shoes. You see there article takes quite a while to load. But eventually, the full content does load in. And you can see there's tons of stuff here. Let's try something a little more specific. Like, what number do I text for resolds? Again, I get the same article. And actually, the answer is in the article, it is 74,637. Let's try something slightly different. How much is a stitching repair? In this case, I actually get a journal entry back about stitching a pair of shoes, but the answer to the question is actually not in this article. It's actually in a different article. So let's jump back to the admin console and switch from keyword search to document search, to our new document search algorithm. This algorithm can find the right document, but it also can serve as the most relevant answer from within the document. So let's turn this on and see how it performs. So we'll start again, resold. And you can see we get the same article back that we got before. So that's great. Now let's try something a little more sophisticated, now that we've turned document search on. Let's try the same query from before of what number do I text for resolds? And as you see, we actually got a direct answer. It scanned the entire document, the entire unstructured document and found the answer and highlighted directly inside this article. This now allows us to get direct answers to searches directly in our search results. So no longer do you have to jump into the article and hunt around for the answer. The document search algorithm services the answer to your questions directly out of the unstructured data. So let's try another search. How much is a stitching repair? And you can see it returns a direct answer of $35. If you remember, not only did keyword search not return an answer, it actually returned the wrong document. So document search is a brand-new algorithm that we launched as part of our Spring '21 Release. So as you can see, with some light setup in up and coming shoe company who prides themselves on their craftsmanship and quality can add natural language, AI-powered search to their website with a few clicks of a button. And so that was a demonstration of the technology that's part of our Spring '21 Release. But that's not where it ends. We have an incredible road map for the rest of the year. We're working on scale, more algorithms, entity-level access control and one-to-one personalization. We have an amazing road map that we're very excited to get out to our customers. Thank you.

Hi, everyone. So now we're going to start our first Q&A session. As a reminder, I'm joined by CEO, Howard Lerman; and Chief Strategy Officer, Marc Ferrentino. We will be taking questions from the audio line, and we'll take questions from those submitted to us through the Q&A chat window.

[Operator Instructions] So now we're going to start with a question from the chat line, and this is for Howard. When do you expect Answers to have a material impact on the business? And why don't you break out Answers from the rest of the business?

Well, thank you, Yuka, and thank you, Marc. Thank you all for coming to our Investor Day today. We're so excited to reveal more of our vision of our company, our vision for the future of search. And our objective here is to tell you about what we're doing and to inspire you with the future of search and the future of Yext. With regard to Answers, Answers is a pretty new business line that we launched just about a year ago. You're going to hear later from Dave Rudnitsky, our President and Chief Revenue Officer; and from Steve Cakebread, our CFO. But as a little preview, we now have 245 Answers transactions closed. This is a pretty good number for about a year. And furthermore, 45% of Yext's deals of our ACV is now multi-product, that is customers that have purchased more than 2 products. And then just one last data point I'd like to leave you all with is that Answers deals on average are almost nearly 3x as great as deals that don't contain Answers. When you sign up with Answers, you typically are a bigger customer, you pay more, so we're really excited about the progress at Answers. And just the big picture here of when this becomes even bigger, let's just say, gosh, we -- there are 5 categories of search: site search, support search, e-com search, app search, workplace search. There are numerous applications of search all throughout the enterprise. Think of you as a consumer, how often you use search in a product, in your car on a phone, everywhere. Everywhere there's a magnifying glass, that's a search box. If that search isn't powered by Google or by Microsoft or Amazon as a consumer service, it's very likely a machine-powered keyword search. And that is our opportunity. We are in the very early innings of this opportunity. We've just gotten started. 245 logos in that first year have signed up with Yext Answers and are paying customers, and we haven't even gone in, we only today announced support search. Think of every help site. Go to any help site. I challenge you all right now, go to help-dot whatever-you-want dot-com, that's a support or help site and try searching in there. Try asking a basic customer support question. Invariably, they all fail. You're not going to get good answers for really basic questions. Try asking a question like, how do I log in? Or try asking a question about how do I reset my password? Try asking a more complicated question. These support sites don't really work that well, and people call up customer support dissatisfied when they can't get the answers that they want. Support search is an enormous opportunity. It's the second category of search, and we only just revealed our document search, our Extractive QA technology today, which is going to let us get into this amazing new category of search. We haven't got into workplace search yet. That's going to happen, we believe, in the second half of this year as we work towards getting features ready to go. And we haven't really gotten into e-commerce search, although we do see a lot of trick -- we see a trickle of demand for it. There are start-ups out there, which are beginning to build Yext Answers into their own product experiences so that they can add natural language search and knowledge graph-powered search into their products right from the get-go. So I've never been more excited and bullish about the opportunity ahead. We believe it's a $30 billion -- our total TAM is now nearly $30 billion by 2024. These 5 categories of search can be extraordinary. Though right now, we are still only in that first category. And we just today announced the set of technologies and features that are going to make it possible for us to get into that second category. Now you also asked, Yuka, the second part of this question, I think, was why don't we break it out? Well, let's talk for a minute about the Answers search platform. So as we grow, our customers purchase a solution from Yext. They buy a platform solution, a search platform solution, in which our core listings franchise fits really well for many customers. It's all powered from the same knowledge graph. Think about it. You put in all the data about stores. You put in all the data or facts about physicians and that same knowledge graph can also power Answers from search, it also can put data into Google and Apple. And so in many ways, it's one single spot to manage all of your smart answers from a single knowledge graph. And our customers don't necessarily break it down like that, and I'll give you a couple of examples. You might close, and this is a hypothetical, a 7-figure deal. And you buy the knowledge graph, and you buy services around it and you buy Answers. We don't really look at it like that. What we do is we build a solution, we build a platform for our client. It's the Answers search platform, and that's what we're really going to be focused on going forward.

Great. Thanks, Howard. Another question from the chat window. For document search, does the enterprise have to put all of its documents into the knowledge graph?

Well, I'm going to turn it over to our wonderful Chief Strategy Officer, Marc Ferrentino, to answer this question.

Thank you. So for document search, what you need to do is you need to put the indexable portions of the documents into the knowledge graph. So for example, if you have a large PDF, well, much of that PDF is made up of header and footer and footnotes and tags and all the sorts of things that are not things that you may want to search. So the part that actually comes into the knowledge graph is the part that you actually want to search on. Another example would be indexing a web page. But you don't bring all the HTML into the knowledge graph. You don't bring the header and the footer into the knowledge graph. You just need to bring the content, the searchable text, the unstructured searchable text into the graph itself.

Great. Next, we'll take a question from the audio line. Operator, can you please open the line for the first person in the queue?

[Audio Gap]

As you look at sort of the 5 categories of search moving forward and the usage you've seen from the, say, the 245 customers already, how is that dictating sort of how quick will you move into some of these other areas of search moving forward? And maybe could you highlight the key features that maybe you need to build out to move beyond support search into the interesting area to me was the workplace search opportunity or e-commerce search?

Ryan, thank you for the questions about the 5 categories search. Site search, we believe, is really big. Every website needs a site search. We have 245 live or customers using site search thus far. We believe we have a huge, enormous runway ahead in that first category of search. A pretty common thing that happens is that customers, when they see our site search, they immediately see it and then they say, "Well, gosh, can we add this to our support site? And gosh, we really love this natural language capability, and we love the analytics that are coming back so quickly, and we love the ability to control all the facts from one knowledge graph across multiple destinations. Can we use the Yext search platform in other areas across our company?" And that's where we began to uncover support search opportunities, which we unveiled the feature set for today. Extractive QA is what makes that possible. Extractive QA allows us to take answers from unstructured text and surface them, which we saw a great demo from Marc earlier. As we get towards -- and let me actually, Ryan, talk for a minute about app search. App Search, today, we announced our developer CLI. It's a Command Line Interface. It makes it much, much, much easier for a developer to add natural language or modern Yext Answers search to their own app. So App search is something and it's a category in which we intend to participate in the first half of this year. And then you specifically asked about workplace search. I'll let Marc talk a little bit about the features that we see that are necessary there. I will say it's a pretty common ask of customers to try to ask us if we can handle this.

Yes. So the things that we need for workplace. So as Howard said, there's sort of an incremental sort of semantic expansion of use cases as the platform -- as we add more functionality to the core platform. It's not specifically about going after a specific search category, it's about having the right functionality in the platform to then enable for those solutions to be built and to be leveraged on top of our platform. So with the addition of the API and CLI, we now have the ability to go after app search with the addition of document search, that opens up support search for us. And now workplace search, for that area, really, the handful of things we need is we need a larger connector library. We need to increase the scale of the overall platform. We're basically row level or document level entity. Row control is a big part of it also, as is also one-to-one personalization. These are all sort of foundational concepts that are on our short-term road map that we hope -- we look forward to showing everybody over the next handful of months.

Great. Thanks, Marc. Next, we'll take our next question from the audio line. Operator, can we take the next question?

The next question comes from Arjun Bhatia of William Blair.

Can you guys hear me okay?

Hi, Arjun. Can hear you great, man.

All right. Perfect. Howard, Marc, good to see you guys. Quick to follow up maybe on Ryan's question, around the additional search categories that you've talked about today. How should we think about what this does from a go-to-market perspective for Yext? Do you still anticipate leading with site surge of the primary landing point for Answers? Or is there an opportunity to -- for one of these other use cases, whether it be e-commerce or play search to become the landing point with -- for customers discovering Answers? And then kind of related to that, is there a difference in buyer persona between site search and some of the additional use cases that you've talked about today? Or do you think there's a centralized kind of buying motion there?

I'm going to answer, Arjun, the second question because I think that will shed light on to the first question, which is when you think about us landing with site search, it really -- we have a big advantage there because we classically have for our customers, the ability to put all their data into a knowledge graph to structure it and to put smart answers into their own site and into Google and Apple and the rest of our listings network. This is a really neat way to get going. Higher persona for support search is still often in the digital customer experience department of a company. It might be a slightly different place within the company, but it's still outside the firewall. And still, ultimately, in the company lives in the customer experience organization. And companies, as you know, spend an enormous amount of time and money and effort to create great customer experiences. Adobe, the leader in DXT, digital experience platforms, claims, I think, an $80-plus billion TAM for that particular market alone. So I realize everyone is excited about workplace search. And so are we. But before we get ahead, I just want to talk about what we have today, which is really big. We have site search, and we now have support search and we have app search. These are 3 big areas that when you're a company, you're not going to necessarily want to have a bunch of different search platforms. Now search unlike listings, and this is a very interesting thing, is not necessarily a winner-take-all market within a given customer and that's how we're able to get in because even when companies sometimes might have a Solr instance or a legacy Lucene instance even powering some old thing internally, we can still come in and get site search. And then over time, we can build from a line of business owner. And that persona could be the CMO, that persona could be from the person who owns the website, that persona could be support search for the person who owns customer support. But then as we build trust, we begin talking more and more to the CIOs and to the CTOs within the company. We're thinking platform wise, horizontally, how to get the best search experiences across not just one line of business, but across the entire enterprise. We've got a really big advantage in that. We don't just offer a point of sort of a point search product, we have a full search solution, which has our listings and our pages in our knowledge graph.

Great. Thank you, Howard. Our next question is going to come again from the audio line. Operator, the next question please.

The next question comes from Tom White of D.A. Davidson.

On the demo of the new site caller feature, it seems like it will help new customers kind of get their knowledge graph up and running and reduce the amount of set up necessary. I'm just curious, like as you look out over the next several years and as the platform evolves, like how much more opportunity do you think there is to even further streamline kind of maybe that onboarding process? Can it get even more seamless, more easy for new customers to kind of get their knowledge graphs to kind of critical mass?

Yes. So Tom, phenomenal question, mainly because I love talking about it. The data -- what you saw was really the first version of a data connector pipeline, not just simply a crawler. The data connector pipeline has the ability to take in different types of data sources. The crawler is just one of the data sources. You can also take an adjacent feed. You can have web host. We can also have the connector library that you see the beginnings of us building out with Magento, with Shopify, with the recent announcement of Zendesk and ServiceNow, Salesforce Service Cloud on the way and many, many more on the way. On the crawler itself, what's kind of neat about it is a lot of this -- is right now, the crawler is able to go to the site, pull certain specific pieces out. But you have to -- there's a little bit of setup that has to happen. You have to point to the pages, you have to sort of point to the attributes inside the page, you got to pull in and you then have to map it into the field. But where this is going is the ability for the crawler to basically look at a page without any setup and simply extract the structured data directly into the graph. And now this is a long-term view. It's not going to happen all at once. It's not like sort of a magic bullet here. But there is, though, is that incrementally, over time, we get more and more intelligent about the structure of information we leverage our natural language capabilities and our AI machine learning capabilities to be able to look at pages on the sand pages. And so you'll start to see things where we can look at an FAQ page or just look at a website and say, "Hey, those pages have FAQs in them. So, let me just pull them in. Let's not bother the user. Or those pages have -- are location pages or those pages are doctor pages. Let me just bring them in." And so that's what you're going to see over the sort of short-term, long-term road map, as we're going to continue to get better and better at that. And then eventually, that's the sort of thing that you could then not just put -- sort of point at a website but even point at all the enterprise data systems and kind of go from there. And so that is a big goal for us because once you have the knowledge graph, as you guys can see, there's so many opportunities, there's so many applications. There's so many ways you can use it. So building that graph is really the first step for our customer success.

Great. Thanks, Marc. Our next question will be from the chat line. It is from Matt Coss from JPMorgan. And the question is, how has the search traffic for map-based searches on Google trended during February and March?

Well, Matt, thank you for the question. I think we said in our Q4 earnings call, that we saw in Q4, on a per location basis, Google Maps traffic declined more than 50% year-over-year. That is a pretty staggering headwind, which was obviously caused by location-based shutdowns for our location-based listings product. Nonetheless, our listings products still grew this year. That is the primary value proposition of listings and listings still grew year-over-year, which I think is a testament, and we believe is a testament to the resiliency and criticality of having listings in Google and other places, even when traffic is severely hampered. We have not yet seen anything that quantitatively suggests that the world is going to come back. I don't -- sorry, that the world is back yet. And I don't think anybody believes for a second that in the future, Google Maps traffic is going to be down 50%. I think that -- we all know that, that feels like a temporary dislocation and that when the world opens up, it will return. But that said, we have not seen anything quantitative. However, it does -- there is a feeling, sitting here on March 17, that things are about to open up. That said, we're still all sitting here on Zoom. We're still all sitting here, in many cases, not going to places that are still closed. My hope, my belief, we all hope that this comes back, and when it does, we will be ready.

Okay. Thanks, Howard. We'll take another question from the chat line. This one is from Elizabeth Elliott from Morgan Stanley. You announced a lot of new features today, especially around Answers. Will Extractive QA or ability to authenticate use cases be included in the current product? Or are these an upsell? How is Answers priced today? And do you expect it to change as you expand into more search areas?

Happy to take that. So the core -- the core philosophy, our core vision, architectural vision of Answers, as Howard has mentioned, and we also showed the demos, is that we are a multi-algorithm-based platform. And so we're not charging anything extra for those algorithms. That is sort of part of the Answers platform itself and as you buy the platform, it doesn't matter which algorithm you use or how much of one algorithm versus other you use, it's all sort of part of that Answers pricing. So they are not, in fact, upsells. The Answers pricing itself will not expand -- it will not change as we expand the use cases. We anticipated -- this was our plan was to expand more broadly as a search platform that could be leveraged across the entire enterprise. And so when we made our changing -- our change to capacity pricing, last year, we had anticipated sort of this expansion into new areas, and this pricing is set up for that. So what we'd like to do is we'd like to see expanded usage of the platform. We'd like to see the platform go from our current model of capacity pricing, where it's based on number of searches and number of entities that you put into knowledge graph. And so as people use it for more use cases inside the organization, the search traffic will increase. The entity counts will increase and thus create upsell -- very organic upsell opportunities for us inside of existing accounts.

Super. Thanks, Marc. We'll take our next question from the audio line. Operator, next question please.

The next question comes from Naved Khan of Truist Securities.

Howard, you talked about the $30 billion or so in TAM. Can you maybe just break that up between listings and the 5 areas of opportunity that you highlighted for Answers? And then with respect to the opportunity with -- on the enterprise side, is there any effort to tie up with some of the tools that are used in the enterprise sort of get easier access to the data sets that we manage you said on top off?

Thanks, Naved. We'll get more into the TAM when we get through with Steve and with DR, David Rudnitsky, a little bit later. We'll talk for a second about the enterprise. Marc talked about the connectors, right? Today's announcement of our crawler is a really important announcement because it now lets us go to a site and drive knowledge from unstructured text, and we can do that with the computer. We're going to get better and better at this. It's a little easier for us to do this than Google. Because Google has to go out and do that for the entire world, we can focus on one limited domain, like an FAQ page and get better and better and better at it since we know what to look for and what to expect. And generally speaking, how to interpret things. The crawler is one way, the first really important way that knowledge, unstructured knowledge can get Dx. But across the enterprise, you're going to need to see connectors, connectors to existing systems. Connectors, I think today, we announced integrations, for example, with Zendesk and with ServiceNow. And those integrations, we can pull in structured knowledge-based articles, those are connectors. So you should expect to see Yext expand the number of connections we have. One of the basic actions we have is with Google My Business. We have a Shopify connection. We can pull in products from Shopify or locations from Google My Business, and now we can pull in support articles from Zendesk and from ServiceNow. So there's a lot of different systems that we want to build these connectors to. This all comes back to Yext's original strategy, which is to be the source of truth, knowledge graph for a company. And could do what Google did to consumers, we want to do that for enterprise. Google brought natural language, knowledge-powered search to consumers, Yext with our connectors, with our knowledge graph technology and with our unique focus on natural language understanding and AI machine-learning understanding is going to bring natural language understanding to the enterprise behind the firewall via workplace search, but outside the firewall via our breakthrough site search engine and via our incredible support search solution, which launched today that can answer questions, like, how do I log in, which other keyword search-based engines simply can't do.

Great. Thanks, Howard. We'll take the next one from the web chat line. We got this one a few times. The Shopify integration looks great, but the majority of enterprises are not using Shopify to power their e-commerce. Will you have similar connectors with Adobe, Magento and other leading CMSs to streamline the knowledge graph population?

Yes. So there's 2 pieces of that. One is, of course, the integrations to CMS, and of course, the other one is integrating into e-commerce background. So actually, Magento is a connector we already have today. So we had actually launched Magento and Shopify back in December. We are going to continue to expand into commerce cloud, into big commerce and other e-commerce back end systems. That's absolutely something that is on the road map, on the short-term road map. As far as on the CMS side, we have our integration with Adobe. We have our integration with WordPress. We are working on integrations with the other leading CMS systems. And when we look at the integrations with the CMS systems, since we have the crawler, we can pull the information in very easily from any website regardless of what it's built in. The integrations with the CMS systems are really around deploying the search results and the tool bar -- and the actual search bar very quickly. It's about making it easy to deploy that specific search bar. While in the case of the e-commerce sites, that's really about making it easy to pull in product information, product catalog information, the information from PIM. And those are areas that all just make it much, much easier to deploy, answers to any site of -- any company of any size.

Super. Thanks, Marc. Another question from the chat window. Why is search best as a separate field on a customer prospects website instead of an integration into existing cap offices, like Zendesk as the Answers their bots provide?

Okay. This is something we think a lot about. We think a lot about how consumers want to interact with knowledge, how consumers want to get answers. And I'm not even sure I think search is best on a separate field on a customer prospects website. I think sometimes it's best. This is a sophisticated answer, so bear with me for a second because we have a point of view here. It's our strategy to have the knowledge and the ability to answer questions to consumers wherever they may be. That could be on Google, that could be an Alexa. That could be on a company's support site, that could be on their site, it could be behind the firewall. It could be accessed through a search box, but it could also be accessed through a chat. Chat is a different type of experience. People engage with chat differently. People tend to -- when they're talking to a chatbot, think they're talking to a person, and they're less likely to be truthful with their question. A lot of times when people chat, they won't reveal what they really want. Whereas when they search, they understand that to be a fundamentally private lookup and will type in something totally different than they're willing to type into a chat bar. And you can see that in the analytics we have, where you can look at the different types of questions people ask search versus people ask chat. We don't take a particular stance on the future of consumer interaction. We only take a stance that the future of consumer interaction will be answers-based, which means there needs to be a knowledge graph and natural language to help people get the best information they want regardless of the way they ask the question.

Thanks, Howard. We are going to take a question from the audio line. Operator, next question please.

The next question comes from Brett Knoblauch of Berenberg Capital Market.

I guess before you launched Answers, you said your kind of TAM was around $10 billion. Then you launched your Answers and you said $20 billion, and now we're at $30 billion. So I guess, can you just help kind of rectify the math behind that calculation? And then one follow-up regarding Zendesk. I guess if your Yext Answers allows you to output the answer some of the search, do you really need Zendesk? Or can this replace what Zendesk is providing you?

Well, first, let's -- we can have a deeper discussion about TAM. I want to be clear, the $30 billion of TAM is 2024. Is that right, Yuka?

Yes.

Right. So there's an implied growth in there. And Steve will be here in just a little bit to help break that down for you, okay? You also asked about Zendesk. We're just providing a search on top of a knowledge base. Zendesk is a full customer success, ticketing suite, ticketing system.

Resolution.

Customer resolution. So it's a pretty different product. What we're saying is, "Hey, if you have a help site powered by Zendesk, a knowledge base with articles, we can now easily pull those articles in. And if you want to add our search to that kind of site using Extractive QA, we're really good at answering those questions."

Okay. Okay.

Yes. I guess if you already have the numbers right there, you don't really need the support that Zendesk is providing, if you are giving the support, right?

The support content all lives in Zendesk as does the ticketing system, which people use. We're taking our search technology and just adding a search box to a help site, whether it's Zendesk, whether it's ServiceNow, right, any customer experience or knowledge base with articles. We can now help get the answer out of those articles. We can mine the answers out so that when the consumer has a question, they can just ask that box and get the answer. But the customer experience systems like Zendesk do a lot more that involves customers and tickets and knowledge base, which we don't do, and we won't do.

Okay. Great. Well, that concludes our first Q&A session. We're going to take a 10-minute break. If you're able to stay with us, we're going to play a great video from our Chief Data Officer, Christian Ward, speaking about building trust in the information age. Thanks.

Hello. My name is Christian Ward, and I'm the Chief Data Officer of Yext. My role at Yext is a bit different than other CDOs in that instead of managing the internal data assets at Yext, I work in market with our customers and partners to help them identify, structure and deliver the right answers to their customers through search. So let's start with the information overload paradox. You've probably heard different names for this paradox in the past, too many choices, TMI, analysis paralysis and the list goes on. The paradox demonstrates that as the amount of information we have access to sky rockets, our ability as humans to process that information actually declines. The problem with information overload is that it disconnects our ability to process our options in a meaningful way, often leading to frustration and indecision. Now don't get me wrong. Being able to search through a universe of information at our fingertips is one of the most incredible advances in human history. But as we will see that massive amount of content has caused a glut upon the customer journey and additionally, led to huge amounts of misinformation online. Unfortunately, this massive explosion in content isn't slowing down. It's actually speeding up. Current estimates from IDC are that content online doubles every 2 years. And that's not including the growth that will come in the next year or 2 from content written and generated by artificial intelligence systems like GPT-3. So for the average consumer or citizen, this massive content explosion is leading us to a poor experience. Now many of us grew up in the age of linear persona marketing. We talk about linear journeys, meaning there's a starting point, a series of steps or touch points, and we track them to a conversion event like a sign-up or a sale of an item. Persona marketing is essentially trying to build different linear paths for different types of people at different stages of their lives. For example, I'm definitely in the middle age, has teenage children, data geek persona category. Typically, marketing funnels have relied on this approach for years. They have a persona they have identified on the left and an outcome sale or conversion on the right. Marketing content, ads, billboards, radio, TV, it's all designed to get the customer on the left to convert on the right. The problem is no one does this anymore. When was the last time you had a question about a product or service? You went to the website of a brand and ended up on a brand's landing page and perfectly followed the step-by-step path to conversion that they laid out for you. Yes, me neither. No one does this. No one does a straight line. They go all over the place. We do this seeking information and the best options. The information age has opened up the opportunity to ask any questions for any brand or business or entity and to explore an extraordinarily unique path before choosing exactly what to buy or engage with. And search has taught us to do this. About 93% of people online begin their digital journey with search. When a consumer starts with search and they land on your website, they're typically greeted with drop-down menus, offers, sign-up forms and generally your interpretation of what they need to know about your brand. Unfortunately, that also means that every single website is a new UI to learn for the visitor. And what typically happens? When the customer can't find what they're looking for easily, they bounce back to Google. And they click on the next link or start a new search or head off to your competition. Now here's unfortunately where the problem of consumers bouncing out of your customer journey becomes a bigger problem. Instead of digital marketers fixing or focusing on a better journey for answering customer questions, they turn to cookies and other tracking technologies to follow you wherever you go. Cookies have long been the consolation price to marketers where they have persistent ads and retargeting meant to bring you back to their website, and you guessed it. They put you right back in that linear persona flow that didn't work the first time. Where does that leave us? How do we enable the customer journey and rebuild trust in this amazing age of information? Well, let's start by examining the word trust. Trust is an interesting word because it actually has 2 very different definitions. The first definition is trust as an emotion. It is literally felt by you. When you think of someone or something you trust, and your brain releases oxytocin. But the second definition of trust is about experiences. It is about the logical set of experiences, whereby trust is earned and gained. See, trust is a loop between these 2 definitions. The more experiences I have with a brand or person where the outcome is accurate and helpful, the more emotional trust I feel toward that brand or person. When that loop is broken, it reminds me of the old adage, fool me once, shame on you, fool me twice, shame on me. So let's talk about brands that have really done an amazing job when it comes to building trust through their digital experiences. Let's start with Google. Look, you might say publicly that you don't trust Google, but the vast majority of us still type in our most intimate questions on a regular basis. For example, my family, maybe like yours, is constantly talking about where we might want to go together, hopefully, once the pandemic is over. I've googled many places trying to find a potential destination. So Google knows this. Now let's say things do open up, and we decide to go where Google knows I'm likely to go with my family. What happens next? Well, next, I'm going to get in the car of a perfect stranger with my family because I also trust Uber. And then I'm probably going to stay in the home of a perfect stranger because I trust Airbnb. Lastly, I'm likely to drop down on the couch in that perfect stranger's home and start to look up audio books because I also trust Amazon. But why? Why do we trust these companies so much? The answer is in their UI. Look at Google, Uber, Airbnb and Amazon. Each one of those platforms is a search first approach. They have spent billions of dollars making absolutely sure that no matter what you ask them, they can provide you with an accurate answer. When we ask questions about a topic, we are sharing both our interests and our intent. It's what we're interested in right now. This is why these companies focus on search to build trust by providing a great experience. But more importantly, the right answers. So where do we start? How do we build an amazing search experience that builds trust, cuts through the noise and doesn't rely on creepy tracking to provide a truly personalized experience? The best way to start is to think about the questions people ask. Every question can be divided into 4 different categories in a 2x2 table. Start with if the question is branded or unbranded for the 2 columns. And then whether or not the question is objective or subjective for the 2 rows. The 4 resultant boxes in the table will really help focus your approach. For example, the question what are marathon sneakers is an unbranded objective question. And there's going to be billions of Wikipedia articles that come back from Google for that type of query. Next, who makes the best marathon sneakers is still unbranded, but that's definitely a subjective question. Here, you're going to find top 10 blog posts and millions of people running content to attract non-branded traffic. Third, a question like, does Nike make a decent marathon sneaker, is now a branded subjective question, and this is where ratings and review sites are going to kick in, probably again, millions of search results, but certainly fewer than the unbranded categories. And lastly, we come to the branded objective question. This will be something like, does Nike sell Marathon sneakers or any other question that can only really be answered by Nike. Because Nike, as the brand, controls what products they sell, what services they provide, what times they're open. Every factual response needed to answer questions like those comes from the brand. The first 3 categories of questions are what I would call classic content strategies, and they're important. But to build the absolute best search experience on your sites, you begin by focusing on the branded objective queries. The ones that only your brand can provide the authoritative answer to. Search is universal, and it's universally understood at this point. Thanks to digital assistants like Alexa and Siri, we now expect to be able to ask questions and far more complex ones at that. I mean, is there any world where you imagine talking less to machines over the next 5 years? Yes, me neither. If the pandemic has taught us anything, it's that this isn't about being digital first anymore. The acceleration of digital transformation has demanded that we start to be our digital best, best-in-class at engaging people in an open, honest conversation where they can trust the answers they are getting. So leverage technologies and solutions that will enable an incredible search experience. When people ask your website questions and they share what they're looking for, you don't need to follow them around with cookies or trackers. You just need to provide them with perfect answers to those questions. By doing this directly on your sites, time and time again, you will build incredible trust. Thank you. [Break]

Welcome back to the Yext 2021 Investor Day. Now I'd like to turn it over to our President and Chief Revenue Officer, David Rudnitsky. David?

Thanks, Yuka. Thank you very much for your time today. I'm excited to tell you about how we've evolved our go-to-market approach and our sales strategy. I want to start by giving you some insight in how we structured our go-to-market strategy. We've grouped the market by 3 customer sizes: enterprise, mid-market and SMB. Our direct sales team sells directly to the end customer. It's focused nearly entirely on enterprise and mid-market opportunities and customers. It's supported in part by our strategic alliances group. Our resellers team is the best way for us to address the SMB customer group. I'll provide more insight on each of these groups, starting with our direct sales team. I'm surrounded by a set of very experienced world-class leaders and professionals, including Brian Distelburger, who's our Co-Founder, President and COO. In North America, Carrie Bosworth is responsible for our CBU Business. Lindsay Johnston in North America is responsible for our Enterprise business. In EMEA, John Buss is our Managing Director for the entire business. And Shimogaki-san is responsible for the entire business in Japan. Mary Fratto Rowe is our worldwide Chief Customer Officer. 5 years ago, we started investing in our sales teams, focusing on our enterprise and mid-market direct salespeople. In that time, we have more than doubled the size of the team from 100 to nearly 250 today. As we noted on the Q4 earnings call, we're planning to increase the team to 255 quota-carrying salespeople in fiscal '22. Though we'll look to accelerate hiring as the economic conditions improve. We have never had a more tenured sales team than we do today. The average tenure, because of our investment in scale, has gone from 18 months to 21 months. Our teams are focused on reducing our sales cycles, growing our pipeline, improving conversion, sharing best practices and accelerating sales. Our direct team works alongside our strategic alliances team. They represent our ecosystem of tech partners, systems integrators and agencies that work with us to create solutions and solve business problems for their client base. These partners often work on digital transformation and other major tech infrastructure projects. By working with these partners, we're able to enter the conversation with the prospective customer at the right time and with the key decision makers, many of whom are in the CIO office. We're very excited about our growing relationships with these great partners. It's still relatively early, but we've already made significant progress with our strategic alliances program. Yext became a charter member of Adobe's premier partner program last year. Since then, we've already closed multiple deals, including some in each major geographic region. We're building a strong pipeline of opportunities in fiscal '22, and we're really excited about our relationship with Adobe. Our partners have become increasingly familiar with our project. We have over 500 Hitchhikers trained at our strategic alliance partners since September. I'll tell you more about Hitchhikers in a moment. Among systems integrators, we wanted to highlight our deepening opportunities with folks like Accenture and Capgemini. We closed deals with each of them this past year. They've welcomed us in with their top tier customers, and we have multimillion-dollar opportunities in our pipeline to go after with them in fiscal '22. The Hitchhikers program is a self-service training platform. It enables our clients and our partners to become Yext product experts, implementers, developers and administrators. There are hours and hours of content on hitchhikers.yext.com to help them on their journey to become Yext experts and drive value, whether for their own companies or for their clients. We have forms, exclusive webinars and office hours and events to build and strengthen the Yext growing Hitchhiker community. This program is only 6 months old, but take a look at the progress. We've had 3,500 plus Hitchhikers trained, 1,900 plus badges earned, and we have 14,000 plus modules completed. They have this huge force multiplier for our internal customer support, our service and our consulting teams because they're capable of implementing, deploying, administering and tuning Yext solutions within our clients and our partners' clients. We can grow more efficiently. We can scale more efficiently and our internal resources are significantly enhanced by the Hitchhiker community. In our SMB world, the best way to reach them is through our resellers. Once upon a time, most of Yext's revenue came from direct sales to SMB customers, but we realized the most efficient way to serve that market was through the reseller channel. Our resellers are experts at selling to and supporting SMB customers, and they're able to bundle our products alongside other software tools that SMBs find helpful. We often have multiyear contracts with some of our largest resellers, located both in North America and Europe. And we expect opportunities within that network to grow, both domestically and internationally over time. What's really exciting is that until recently, our resellers only sold listings. But in January, we announced we're enabling select resellers to sell Answers. And we expect that to be a nice addition to growth for the reseller program. If you look at our platform, it's innovated over time. And now we have a broad set of technologies and features on the platform. The Knowledge Graph is the foundation of the platform, it's the foundation for all our offerings and we build solutions on top of it. We're able to take these capabilities to create a diverse set of solutions and solve customers' unique problems. As we continue to increase the capabilities of the platform, we'll be able to build even more search solutions on top of it. We're solving our customers' problems. So increasingly, they're buying multiple solutions to address their digital transformation strategies and goals. As a result of that, we've seen a steady growth in multiproduct customers. An increasing portion of our ARR is from customers who purchase solutions from 2 or more of our major products, Answers, Listings and Pages. 4 years ago, only 23% of our ARR came from multiproduct customers. Today, 45% of our ARR is for multiproduct customers. The addition of Answers in late fiscal 2020 and accelerated this dynamic. There's been a steady increase of customers with $100,000 more of ARR. It's really attributable to 2 things: the ability to cross-sell additional products; and the ability then to go back and upsell capacity, upsell usage, upsell the Knowledge Graph data to our customers. Our Land with Answers sales motion quickly drives value in the form of a few things: one, there's higher conversion and lower support costs; two, there's clear and measurable ROI, supported by our analytics; three, it's intuitive, it's easily explainable, there's an improvement you can see immediately from our prospective customers' existing site search solution, there's meaningful momentum in both customers and search volume, and we expect to see continued strong growth in FY '22. In FY '21, new logo deals with Answers were nearly 3x larger than the average deal size of new wins without Answers, 3x larger. We think this suggests the power and possibility of our Land with Answers sales motion and our ability to use our whole platform to solve customers' problems. We also wanted to highlight for you the opportunity for us to upsell with additional capacity on Answers. Recall, we're a capacity based pricing model with Answers. Our customers buy based on an amount of search volume over the course of their contract. In this example that you see, one of our large CPG customers initially purchased amount they felt was reasonable but quickly discovered that their search volumes outstrip the purchase capacity by a meaningful amount. And so 6 months into the contract, the customer increases is purchase for a capacity amount that met its needs. We see capacity upsell opportunities with a number of customers. And believe as customers see the ROI from having a great site search experience that we'll have continued opportunities to upsell in this way. Answers also opens up opportunities in verticals where we hadn't seen as much traction with Listings and Pages alone. In the past, we primarily serve 6 major industries: retail, food services, hospitality, financial services, health care and communications. With Answers, we've seen increasing opportunity in a number of other verticals, including infotech, CPG, government, higher education and digitally native e-commerce companies. It's really encompassing every type of business. For our initial sales of Answers in 2021, we saw a notable shift towards financial services, health care and communications companies. These were areas which performed relatively better during the pandemic and away from retail and food services, which were more impacted. We've told you in previous earnings calls that retail and food services was 25% to 30% of ARR. We wanted to give you some more insight on our vertical mix today. In the past, we had more significant exposure to those areas, particularly the retail. We've seen retail, as a percentage of ARR, decline by 8 points over the past 4 years. As we've diversified our vertical mix away from retail and towards areas like health care, while the financial service industry continues to be a key and strong market for us. As we move forward with Answers, we do expect our opportunities to broaden with verticals in that other category like government, CPG and digitally native. These are verticals that open their businesses post COVID, we expect to see them start to grow again. Finally, I wanted to show you our strong penetration across a number of industries with well known, world-class, top global brands. Our platform has become a critical part of their digital strategies and transformations. It has also improved their customers' experiences. What's exciting is that we're just getting started. With that, I'll turn the presentation over to Steve. Thank you.

Hello, everyone. And today, I want to talk about 3 things: our growth, our key business drivers and our outlook that we have going forward. So regarding growth. As you can see, we've had strong revenue growth over the last 5 or 6 years. Even our CAGR with FY '21 still comes up in the 30s. And we're still focused on growing this business consistently over time with new products, new opportunities and new channels. You'll see as well that the TAM has been growing. And that's a result of the products that you saw today as well as our constant impact on trying to expand our TAM through customers, industry segments, new products and our international expansion. With that as well, we've had good strong cohort analysis and growth in cohorts over time as well. So we're happy that our customers remain with us and enjoy our products. I think that's because we keep adding new features and functions to our products as well. And we keep expanding the opportunity our customers have to explore the search information they get out of our solution. At the same time, net retention this year has been challenged as we've talked numerous times on our earnings call, that upsells have been challenged for macroeconomic reasons but we believe that as we bring new products back into the market, like you've seen that Marc and Howard brought today and economic recovery comes along and our expansion of our businesses, we'll see that retention get back to the 110% or on average where we want to be. I'm most excited about our international expansion. It's 20% of our revenues, and that's just in the European region primarily. So we have good growth opportunities there. And being at 20% gives us a lot of room to go forward. Europe has been a powerhouse. Even in the macroeconomic times that we've struggled with, they've been able to keep their growth rates going and become a larger part of our revenue going forward. Another opportunity has been our gross margins. As you can see, we've improved them 10% over the last 5 years, predominantly because of scaling with our publishers, but also more efficiencies in our delivery mechanisms as well. And our guide is 75% to 80% gross margin. And I suspect that we're going to continue to drive to get that to the higher end of our guide, if not a little higher beyond that. Now let's talk about key business drivers, a very important area for us going forward. As you know, we're focused on 2 very broad areas: driving growth and driving productivity. The main driver of our growth is product innovation. Over a decade, every year, there's been new features and products coming out. And this year is no exception. What you've seen Howard and Marc deliver is wonderful. And that's going to carry on. And it's a big driver for growth because it brings us into new TAMs, it gives us additional products to upsell to our customers and helps us enter new markets like search, a major new market for us and a great opportunity in a number of areas that we're just starting to touch. With this growth in products, allows us a chance to expand our delivery mechanisms. We've obviously always been a direct sales and reseller. And you've seen we've made big investments in the direct sales. But we've also started to invest in services. So we make our customers successful. And we get the feedback that we need to make our products more successful over time. But in addition, we've added another channel this year, strategic alliance and partnerships. Another opportunity to grow the business, another opportunity to enter new markets and another exciting chance to meet new customers in these particular arenas around strategic alliances. As we've said before, international has been a key part of our growth so far this year, and it will continue to be part of our growth opportunities going forward. In the future, we expect international revenues to be at least 50% of our business. That gives us a lot of headroom to grow our business simply internationally. And it's not just in the France and Germanys and U.K.s and the Europe market. It's also in new markets, particularly in Asia, where we've had limited operations primarily in Japan. But as we can get back to countries, we'll start to see us expand into additional countries as well. So international expansion is going to be a big part of our growth. And like I said, I expect it to be about 50% of our business going forward. Onto the productivity drivers. We've talked a lot about sales and marketing. We've made good progress over the last 1.5 years in improving sales and marketing through changes in how we sell. We don't send a bunch of people to an office anymore. We can send 1 or 2 and still be just as effective to the tenure of our sales organization, as David talked about earlier, and to the processes that we have and the new products that we have to make our sales executives more efficient. But it's not just sales. We're working on all kinds of operational efficiencies throughout the company. New processes, new systems that are more integrated and more efficient and new ways to address our customer and our own processes. One example is a continued improvement in G&A, but it's across the board where we're making investments to make ourselves more efficient over time, so we can scale with the growth we anticipate. Now let's talk about our outlook and where we're going. As you know, there's 3 channels that we serve: direct sales, which is mid-market and enterprise; reseller, which is all business; and then our services, which is directed at our customer. These combinations are going to help us continue to grow our business, and we've got to focus in each of them now that we need to have to continue to help us grow across the spectrum of customers that we serve. You can see here our investment over the last couple of years, while it looked high, has really paid off. Even in a pandemic year, direct grew 25% year-over-year in fiscal year '21. So our investments are making sense, and they're proving themselves to be successful over time. And we expect direct sales to carry on that growth rate. With regards to resellers, yes, obviously, it's a challenged year for small businesses around the world. And we have resellers in Europe and North America predominantly. So all of them suffered through lockdowns and lockups, et cetera. We're going to treat the reseller business fairly conservatively because we think it will take a little longer to come back. But it's still a major part of our business. We're excited about the opportunities we have. As you know, we introduced answers into this channel, and we expect that will start to bring the growth back over time. But we're going to look at this realistically as the economy start to open up in the near term. And then lastly, as we said, we've invested in services, and you've seen that channel growth tremendously. It's a big part of where we want to be with our customers. And like I said, the feedback you get is incredible. So I think we've made the right investments in the direct channel to help that grow. We're clearly continuing to support our reseller business, and we started our services business, and all 3 of those, we expect to grow in the future. When you look at our model then in the near term, our direct channel should continue to grow in the 25% plus range over time. Again, we'll be adding quota-carrying headcount as we see the economy turn around. We'll be supporting the resellers with roughly a 5% growth through new Answers products and opportunities, and the services business is just getting started. Now we don't expect to have services be a large part of this business. But like any SaaS company, services will be 10% to 15% of our revenue. So this growth is important and putting it in place to serve our customers and make them more successful is critical over the long term. Our longer-term model has us with gross margins of 75% to 80%. As I said before, we're right dead set in the middle of that, and we expect to continue to improve those gross margins over time. But we like that range for now. Sales and marketing, 30% to 35%. That's a goal. We're on that trajectory to get there. But again, these are longer-term goals, but we believe that we can be much more efficient in sales and marketing over time. R&D, 10% to 15%. Now this is where I think you might see us spend a little bit more in R&D and make sure as we get new products that we're still consistently driving product innovation over time. But there's a trade-off between having a lot of R&D people and having an effective R&D team. And we think we have a very effective R&D team. So they can do the job. They're bringing out a huge amount of new products this year and I'm sure they'll carry on in that area. But we're also looking to make sure that we're spending into our R&D over the foreseeable future. And then G&A goes to about 10% longer-term and gets huge benefits from the systems and processes that we put in place and becomes more scalable as we grow our business. That will result in operating margins of around 20%, and it's fairly traditional for a software company. We've had 3 tenets that we've been running this company by: grow revenue, get to operating cash flow breakeven and get to non-GAAP net income breakeven. While I'm here to say and you heard in Q4, for the full fiscal year, we were operating cash flow breakeven for fiscal year '21. And as you can see on this chart, we've been cash flow breakeven or better, 2 out of the last 3 years. What we said on the call, and we believe this, we will operate at cash flow breakeven for fiscal year '22 and beyond. The reason why is we will spend into our growth so long as we stay in the cash flow breakeven range. So we want to make sure that we are breakeven, but we're not going to short change our growth to make that goal. We'll make sure we're growing, but we're going to grow smart. We're going to put sales reps in territories where we have leads already. We're going to make sure our processes support the new businesses that we get into. If you do all of that, we should be able to run this company at cash flow breakeven and beyond for a long time. Now with regards to non-GAAP net income, of course, that's always been a goal. We've been close. We've decided and we've looked at our business and believe that macroeconomic things change, we will get to non-GAAP net income breakeven by fiscal year '23 on a full year basis, and we can remain there as well on a full year basis. So that's our next goal. We've got revenue growth, and we're going to continue to push on our revenue growth. We've made our cash flow breakeven goals, and we're going to keep those. And the last one to follow is non-GAAP net income breakeven, and we think we can get there by fiscal year '23 and beyond. With that, I'll be glad to take questions with the rest of the Yext management team joining me. Thank you all very much for attending today.

Great. We're going to start our second Q&A session now. Joining me are CEO, Howard Lerman; President and Chief Revenue Officer, David Rudnitsky; Chief Strategy Officer, Marc Ferrentino; and Chief Financial Officer, Steve Cakebread. As a reminder, we're taking questions both from the audio line and from the Q&A chat window. [Operator Instructions] So we'll start the session with a question from the chat window. Can you double-click on your $30 billion TAM? Does that include listings? Can that inflect your growth to grow?

Thanks, Yuka. Let me take that question. I know a number of people have asked that. As you remember, we started a company with a $10 billion TAM that looked at our Listings business on a global basis. We've added Answers. And when we did that, we expanded our TAM, and our estimate's about $20 billion. And as Howard said, we're estimating the growth in 3 to 4 years to bring that TAM to $30 billion. But as I said in my remarks and I continue to believe, as we can bring out new products, as we expand our market entry into additional segments, you're going to see that TAM get bigger. So this is just the start of the business that we see going forward. And as you know, TAMs are hard to estimate but this is our best guess going forward right now. Listings at about $10 billion, Answers at about $20 billion, growing to $30 billion and our ability to expand new products will continue to expand that TAM.

Great. Thanks, Steve. Our next question will be from the audio line. Operator, can you please take the first question.

Our next question comes from Ryan MacDonald with Needham & Company.

Another question from me here. I guess this one is for Steve. As you outlined the medium-term financial outlook and the growth targets there, what assumptions are you making to get to those targets in terms of a recovery in the listings business, continued momentum from Answers and some of the additional product functionality that you announced today in terms of the other areas of search?

Ryan, thank you. That's a good question. We know that the SaaS model comes up slowly. It goes down slow in difficult times and it grows slowly. So we've taken into consideration, clearly, future expansion of Listings. As we said on our Q4 call, though, we're not really able to call the timing of when things come back. And as Howard really described well, we all feel it, but we don't see it yet. But we do believe that over that 2- to 3-year period, you're going to see this company, again, with new products, international expansion and the recovery of Listings business get back into the 20% growth range at the end of that period.

Okay. Great. Thank you. Operator, we'll take another question from the audio line.

Our next question comes from Naved Khan with Truist Securities.

Yes. Maybe just amongst the different opportunities that you guys outlined for Answers, if I have to think 3 to 5 years out, which is the one that makes you most excited about in terms of potential for growth?

Naved, that's like making me pick between my kids. I couldn't be more excited about site search, I couldn't be more excited about support search. Personally, I am over the moon about App search because I have seen start-ups put our natural language search into their app to create a magic experience. I'm not sure if that's going to be the biggest revenue, but oh, my goodness, the ability to see a well-funded start-up put our technology, bake it into their product from the ground up and provide natural language with Knowledge Graph, the creativity, it's just mind-boggling. I think as these companies launch, you will see Yext-powered search in start-ups. So I guess maybe I am picking a child here. Though that might not be the biggest TAM, personally, I am just really excited about app search, site search, support search, workplace search, or even talk about e-commerce search. They're all -- they've all got a ton of potential. What do you guys think? What most excites to you, Dave?

Well, I think for me, it opens up new opportunities, and we have continued conversations. So as we gain interest in site search and very quickly support search, all of a sudden new opportunities get created. And the idea that we leverage the exact same Knowledge Graph gives us a huge advantage. So I'm super excited about it because I think those other 3 are going to exponentially take off once we get going on them.

Yes. For me, I think really the thing I'm most excited about is the platform, the platform itself, the platform conversation. Search is one of those really interesting things where it is almost more of a horizontal. It is almost more of a piece of infrastructure at most organizations. And for us to become that critical -- mission-critical piece of infrastructure that can be leveraged across a number of use cases, a number of solutions. And all the solutions we talked about are all going to be possible on the platform. But there's so many other variations of those solutions, so many other opportunities and so many other use cases that exist out there when you have a generalized platform. And so the opportunity that I'm most excited about is to become that platform, that search platform for every single organization out there.

Naved, I think what you saw in that answer is a little bit of our -- each of our own individual personalities. I love the start-up stuff, Dave wants to sell the most solutions, and Marc is talking about the technology. We're excited about all of them.

Yes. No, it's still great color, and I appreciate that answer. I have a quick follow-up for Steve. Steve, if I have to think about the opportunity with making alliances and through partnerships, I think you said 15% of revenues. Is that something that we can get to over the next 3 to 5 years? How should we think about it? And then with respect to resellers, can you maybe just size that for us, how big the exposure is? And you said Answers can drive 5% incremental growth. But what's the baseline growth we should assume there?

Sure. So a couple of questions there, Naved. One, as we had said, services would get to 10% to 15% of our business. I'm not sure strategic alliances. They could certainly get bigger, but they're going to be based on the fact that Dave's direct enterprise sales team is going to help that as well. So that will all be kind of into the direct enterprise sales numbers as much as anything. With regards to our partner business or our reseller business, it's always been a big part of our business. And as you know, we've always broken out the growth rates of direct enterprise. What we do believe, though, is they've got great handle on our small business. Now the last year has clearly been tough on the small business segment, and our resellers have done a great job of hanging in there, the slower growth expectations. So we're really baked around a conservative view that small business is going to be tough to come back in the near term. But we've had and been very successful with a number of our resellers. They've been with us for multiple years. They commit on contracts, and so we're excited about that. We are and did just introduce Answers into some of those reseller channels. So that will help the growth as well. But I think we've got to be a little bit cautious about how quick small business can recover, and that's their target segment.

Great. Thanks, Naved. We'll take the next question from the chat line. It is from Matt Coss from JPMorgan. And he asks for Steve, how will you ensure your gross margin improves over time as you expand your professional services? Also, is there an approximate time frame on your long-term target model?

Yes. So long-term target model is a long term. But typically, you're going to look out 10 to 15 years. The way to measure the time frame on the long-term model is when you see our growth rates start to drip into the low 20s and high teens, you'll know we've started to look at that long-term model. And my belief is as growth rates slow, earnings and cash flow grow. So no real-time frame at that point, but look at that as kind of that challenge. With regards to margin growth, we're always looking for efficiencies. And I said in my presentation, it's not just a one-and-done here. We're going to continue to sort out and search for margin efficiencies not at the expense of growth, however. But as we grow, we need to scale. So I think you'll continue to see scale in the margin numbers that we get over time.

Great. Thanks, Steve. The next question we will take from the audio line.

The next question comes from Arjun Bhatia with William Blair.

This is probably for Dave. I wanted to maybe touch on the relationships that you have with the GSIs a little bit more. I know you mentioned Accenture and Capgemini, but we'd love to hear the nature of those relationships, the work that they're doing with Yext? Are they actually doing the implementations? Are they helping customers build site search solutions more on the Listing side? If you could just maybe give us a little bit more color on the existing relationships, and how large you think the GSI channel can get, that would be very helpful.

Sure. So Arjun, thanks for the question. I think we're just getting started with the GSIs like Accenture, Deloitte and Cap. I can tell you that we've spent a lot of time this past year getting ourselves aligned from executive level on down, and the field on up, and we've started to get a tremendous amount of momentum. So we've closed in every one of our major regions. We've closed significant sized deals as a result of working with the teams on the ground for Cap, Deloitte and Accenture. The exciting part about it is, is that they're starting to look at us as a platform. And when you start to think about a platform, that translates into solutions, solutions to translate into much larger opportunities. And that's really where we're starting to head up. A perfect example, there were 2 deals that we closed in Q4, one with a major financial services firm, one with a health care. And we were brought in by the 2 different partners to participate in their day with that client as filling in a solution that they were building for them turned out to be a terrific opportunity for Yext. I've seen this movie before. I think we're right at that point now with the platform. That is a perfect time for us to continue to partner with them and continue to accelerate that.

All right. Thank you, Arjun. Next question we will take from the web. The question is, Answers is a usage-based pricing model, and you are targeting a land and expand strategy. You also expect Listings to come back at some point in the near future. Could NRR be well above your 110% target if this plays out as expected?

Well, from my standpoint, it could very well. But also there's a lot of moving pieces in this business for the next couple of years. So we're just going to be back to what we think are normal levels over time. And then, yes, clearly, if we're more successful over time, the numbers can go up. But right now, our goal is to get back to where we were, and then we'll build on that base. And that's going to take a little bit of time given the macroeconomic circumstances we're at.

Great. Thanks, Steve. Next question, again from the chat window. Dave, do you need to make further changes to your sales motion as you expand into these other areas of search? Sales enablement and clarity of message is important to ensure execution without sales interruption.

All right. So I'll take it in. First part of it is, as Howard talked about earlier, our sales motion whether we're selling site search or as we evolve into support search, is somewhat similar, a little bit of a different target audience, but I think our sales motion, with a little bit of tuning, we could repurpose that playbook very quickly. There's no question about it, though. We're going to have to evolve our playbook as we get into app search, e-commerce search, workplace search. Different set of buyers, different set of requirements. At that point, you start to look at scalability, reliability, security. We are starting to evolve with those. So as part of our planning for that is that we have now set up a team, we call it the search innovation team, very much like any other enterprise company. Whether you call it an overlay, a co-prime, a product specialist, we have a team on the ground with us now as we start to dig in deeper to these other use cases. So we are preparing ourselves for that. I feel very comfortable the playbook will evolve over time. As far as enablement clarity of message, there's no question about it. That we really started to have some terrific momentum with Answers. There's not going to be an interruption. We constantly fine-tune our sales enablement. We're pretty diligent about it every couple of weeks. We have updates to our sales enablement, our go-to-market strategy, the tactics, et cetera, for our sales team. I feel really comfortable we're headed with it. But clearly, there won't be any interruption in terms of our sales motion.

I'd like to add something to this question. I was really glad to hear you use the word playbook. You used the word playbook, I think, 3 times in your answer. I just finished Ask Your Developer by Jeff Lawson, Twilio Founder and CEO. A wonderful book. In many ways, Yext or what we're becoming is the ability to just sort of add search as a service to your application or your site or support site or your app. And I was not surprised, but I was delighted to see on Page 212 in the book, Jeff Lawson's [ on-site ] playbook. You didn't even know that was going to happen.

I have never met Jeff.

Never met. So when Dave talks playbook, I really do believe you with that. And I'm really excited to partner with you as our President and CRO to make something big happen.

Yes. I mean, listen, if you really want to evolve, and you could see the rapid pace of innovation when you look to what Marc presented to us, the only way you can address that innovation and take advantage of it is to evolve. How do you evolve? It's got to be pretty systematic. It's got to be repeatable. It's got to be scalable. I don't know any other way than to have it in a playbook, continue to evolve it. Because as we start to grow our company, we've all got to be on the same page and take advantage of that innovation.

Great. Thanks, Dave. Next question from the chat window. This is from Elizabeth Elliott at Morgan Stanley. And she asks, while it is early, how should we think about the expansion opportunity from opening Answers to channel partners? What do you need to see to expand it more broadly from select partners today?

Howard, do you want to take that or you want me to take it?

I'll take that. Remember, our channel partners are selling to small businesses. That is the primary way that Yext addresses small businesses. We definitely see an opportunity there. Every small business has a website, every small business needs a presence, they need to be able to answer questions. And in addition to that, what we have seen our channel partners using our technology on their own sites, many of them around the world, believe it or not, come from a legacy yellow pages background and handle consumer search shock. They have the sort of keyword based what, where search engines in a random European country or Asian country, for example. And we have the opportunity to bring natural language to them if it's in 1 of the 7 languages that we support in German and French and Spanish and Italian in English -- sorry, 6 languages and Japanese. And so we both can sell them as an enterprise where they uniquely have a local or location search consumer experience. And in addition, they can sell-through to their SMBs all in the same kind of capacity contract.

[Operator Instructions] We'll take another question from the chat window, and that question is, how does Yext capacity-based pricing for workplace search compare to elastic search pricing method? How does the competitive market evolve between the 2 of you? It seems like you have coopetition.

Marc, do you want to take that?

Yes, sure. So as far as pricing, the pricing levels are slightly different in that the elastic search model is more based on a number of servers. It's based on an infrastructure model versus ours, it's just based on straight up usage, number of searches and the amount of data that we're storing. They're similar and they sort of -- they both get to a similar number or near a similar number, but there's sort of a different methods to get there. As far as cooperation over the years, I think that each organization, we're sort of focused on slightly different things. Elastic is sort of the original company that made keyword search a popular and they use it not just for workplace search, but they also use it for APM, they also use it for security and other areas. So the -- so it's like keyword search, while we believe that from a consumer standpoint, it's not the best experience, but keyword search does have applicability in other areas of computer science and other areas of technology. And that's where you see Elastic has sort of expanded into. There is a world where someone is maybe potentially leveraging Elastic for some of their search, but then leveraging Yext natural language AI as to basically bring those 2 things together and create a single unified search experience for the customer that's leveraging the best of both technologies, that is definitely something that we're starting to see more and more of. We're starting to hear more and more of it. And that's one of the reasons why we released our SDK. Our APIs in the spring release, was to enable that to happen in a much bigger way going forward.

Great. Thanks, Marc. Next question from the chat is from Naved Khan from Truist Securities. And the question is, has rep retention been an issue given the slowdown in sales growth? That's for you, Dave.

Sure. Hey, Naved, it hasn't. And we've been pretty fortunate that even with a change in sales growth, our retention rates and our churn is no different than any SaaS company of equal size to us.

Great. Okay. Thanks, Dave. Next question. This is a long one. Let me just read it all out. What are you doing to teach and build an ecosystem of SI in university developer programs, in undergrad business schools and grad business schools and computer science departments and more broadly? How can Yext become a verb where people list knowing how to use it on their resumes and ask for it in the workplaces, if it's not already there, or ask for more of Yext products?

Well, first off, whoever asked this question, please send me an e-mail at [email protected]. We'd love to hire you to help us build this program out. We are looking to build out our Hitchhiker program, and we have a number of open positions. And clearly, you have a ton of ideas, and you've articulated what we would like to do. Hitchhiker, our Hitchhiker program is something that we're really excited about. We've evolved it a lot. We've now got developer personas and better documentation. We want folks to be able to -- instead of having to show up with your developer to build a search engine from scratch or on top of it, we've seen powered search like Solr or Elastic, our vision is to enable developers to easily add search to any application with a low-code type of environment. So you're going to see us continue to build out our Hitchhiker program, continue to expand on the developer-first nature of our Hitchhiker program and my hope to the questioner here is that what the vision you've articulated is something that, over time, as Yext becomes the company that brings natural language search to the enterprise, that you will see people talking about their Yext capabilities on their resumes, too. Send me an e-mail, let's talk about it.

Thanks, Howard. All right. Next, we will take a question from the audio line.

Our next question comes from Ryan MacDonald with Needham & Company.

I've got 2 questions. First one for Dave. You talked a lot about sort of sales capacity and maturation of reps. And I think Yext did a really great job of sort of keeping the bench full during the pandemic. How are you feeling about sales coverage, sales capacity as things start to open up? And what would you need to see to move from, as you said, sort of the 250 today to 255 and above? I guess I'm asking how are you feeling about your ability to service increasing demand with what you've already got in the sales organization?

So, Ryan, there's 2 things. One, we're going to get to 255. We're on track to get to that 255. I think because my friend, Steve Cakebread, have been given permission as the business improves, more so the economic conditions, I can continue to hire. And one thing I learned a long time ago at Salesforce, you got to fill your capacity when you have it. And we will fill that capacity as the economy starts to open up, and we see things turn. But we'll get to that 255 very quickly. The other part of your question, what changes a little bit is we're just going to evolve in terms of the skill set and the type of folks we hire. As we start to look at opening up that platform for search, it will take yet another type of skill set, and that just evolves over time. And it's like a sports analogy, if you hire a different type of athlete as the game starts to change over time, and that's what we plan on doing.

Excellent. And then my second question is probably a combination for Howard and Marc. As you move into these other areas of search, do you expect that you'll need to do the same sort of customer education that you had to do with Answers initially? And how do you think the best way to do that? Is it another 90-day free trial? Or how are you starting to think about that?

A lot of the education we had to do was simply educate our customers and prospects and generate awareness of this idea of natural language search. You don't -- when you run a search, you see the box, you type something in. You don't know what's going to come out of it. If you're using Google, you've been trained that you're likely to get a pretty good result, you're likely to get an answer. If you're using an internal app, you probably kind of know how it works through trial and error. So we've had to educate folks on the fact that, a, there is a difference in search, keywords versus Answer search. We've also had -- and we're still in the process of doing that. And furthermore, we've had to educate companies and prospects and customers on the idea that Yext can actually, believe it or not, bring a Google-like search to your website, to your app, anywhere you want a Google, you can have that, you can purchase that, and it's really easy. You can turn it on and it's faster and it's better and it's cheaper than buying keyword search and building your own keyword search engine and trying to do it that way. So that has been a process. However, I don't -- once we have educated some of the executives within a company, that there's a different kind of search and that we are and have a better, faster, cheaper version of it that they can just turn on. Those executives, their minds start to go and they see other applications for where we can go. So I don't see the need to -- this will not be as hard to launch new search categories as it was to put the company into the first search category on top of our listings and Knowledge Graph platform. That said, there's still a lot of education, a lot of opportunity to put out there. And I'd just like to note, Ryan, you saw us over the past year do that, re-educate folks on our capability and prospects on our capabilities, and at the same time, do so while getting far more efficient. Look at the sales and marketing efficiency last year on a percentage basis over the previous year. I think in Q4, we saw a 7% improvement year-over-year. So not only did we do something really hard, we did it while introducing a set of tactics that were more efficient. So kind of going forward, I think once we've landed with one line of business with search, we'll be able to expand with other lines of business in search and hopefully, bring a Google like experience everywhere across the enterprise in and outside of the firewall.

Great. Thanks, Howard. This next one is from the chat window, and it is for Steve. Steve, the question is, how much growth capital do you envision Yext to require to fund growth during the next 3 years? Will you be looking to capital markets to fund it?

Yes. The answer is not at this time. I mean, we are well positioned. Our balance sheet is strong. As you saw, we were cash flow breakeven this year, full year. As I said, we believe we'll be breakeven going forward on a cash -- operating cash basis. So at this point in time, we have plenty of operating cash, and we're really excited about making the investments to continue to grow the business. So no capital markets. But you never say never. Things change and things -- opportunities come for us. But in the near term, no going to the markets at this point in time.

Great. Thanks, Steve. All right. We'll take the next question from the audio line.

The next question comes from Naved Khan with Truist Securities.

Just had a follow-up. On the 245 or so customers that you have for Answers, how much of those are new versus people you might have upsold the existing Listings customer into?

It's a combination of both. I think the best way to get a sense of kind of like what the percentages are and who we're signing up is to follow us on Twitter. Our handle is @yext. And practically every day, we tweet about a new Answers experience that's gone live on a brand. We don't -- we can't do all of them because not everyone lets us do it. But there are 245 at least Yext search boxes out there on websites. And each day, we tweet kind of a new one that goes live. And you can see, Naved, the different types of industries we've gone into. Dave had a great slide in his presentation that showed our classic verticals, finance, food, telco, retail. And then we showed some of the new verticals that Answers allows us to sell into, infotech, government, CPG. These are the kinds of new customers we can bring on. And given the fact that Yext has never sold into those types of industries before, when you see that kind of logo show up, you can almost be sure that, that was a new logo. The State of New Jersey doesn't buy Listings from Yext, but they did buy Answers from Yext. Companies like Slack don't buy Listings from Yext, but they do use us for Answers. So these are the kinds of new companies we can go into. And when you see those show up on our daily tweets at Yext, that's where you can kind of get a sense of the kinds of companies that are coming through. We'd like to just put them out there because they're just needing their fun experiences on the search side.

Yes. Understood. And maybe just a quick follow-up on sort of the opportunities between enterprise versus the midsize. I think just before the pandemic hit us, you guys had sort of ramped up the midsize. So as we sort of come out of this and things start to pick up, is it clear that, that midsize is going to be like a big driver or also things have changed in between, right? Because Answers seems a lot more relevant to a lot more people. So how should we think about sort of the -- which one is going to be more of the engine as things recover?

Naved, if you've ever been to Yankee Stadium in the seventh inning stretch, they always have which subway is going to win. And they have the -- I don't even know the red line and the blue line and the green line, and there's a race. And everyone is trying to figure out whether EMEA is the biggest, or the EBU is the biggest, or the CBU is the biggest. Quite frankly, it's not even obvious to us. We believe there are big opportunities in every segment. If you look in the midsize segment, for example, what we call our CBU, the new economy companies in there are extraordinary. We sign up Bitcoin companies. I see us talking to trading applications, cryptocurrency applications. We can't serve them, but there's a ton of CBD companies that contact us and want to work with us. There's a lot of questions out there in telehealth companies. So there's all these new categories. And by the way, when you look at app search, that is from well-funded tech start-ups that are looking to build search into their products. Those are not enterprises yet, but they're backed in big ways and some of them have aspirations to become enterprise. So that said, at the enterprise level, you see companies that need support search and workplace search. The opportunities for search are really big, really big. Search is not a winner-take-all market. We believe we can be very competitive in all 5 categories of search and offer enterprise, or a midsized company, or a company in Europe or in Japan a search platform to serve any solution that they need, and also put their smart answers in Google and Apple and Facebook with our list.

Our next question is from the chat window. It's from Rohit Kulkarni from MKM Partners. And his question is, over the past 12 months, you have talked about federal, state and local government agencies. What's the latest thinking about the opportunity in this segment? And do you need to hire more or different types of sales reps specifically for this segment?

Yes. So we are, Rohit. And we've thought a lot about it as well because the opportunity in federal in particular, is enormous. We've had success in both federal -- actually federal, state and local. We are in the process of hiring someone to run that group for us that will have reps that have domain expertise. That is a very different market. As you know, a lot of it is based on long-term relationships, not only with customers but with agencies and with affiliates that can sell into those agencies. And it's very much on our immediate product -- I should say, our immediate go-to-market strategy. So it's very timely of your question.

Great. All right. Thanks, Dave. A second question from Rohit at MKM Partners. On the incremental TAM from new search categories, do you believe that early wins in these search categories would come at the expense of search advertising budgets at those companies?

Yes. So I -- a lot of the new search categories we're talking about are going to come from brand-new budgets, brand new areas. So for example, when we talk about support search, that's coming from the customer experience, your customer support budget. Workplace search is coming typically from IT or even a human resources budget. E-commerce is coming from a completely different budget. So these new search categories are really amazing for us because they bring us outside of the sort of classic consumer marketing world that we have been in traditionally as a company. While we'll still continue to obviously move the ball forward in that category in the marketing category, these new ones are coming from really different budgets and completely different parts of the company.

Great. Thanks, Marc. Another question from the chat window. Can you give a sense of what percentage of your product or R&D organization is now focused on building out Answers functionality versus Listings? How are you managing the potential disruption caused by the shift in focus? Well, you need to spend more, your target model implies you would.

Marc, why don't you take that? But also, maybe let's talk for just a second about our ML and AI capabilities as well.

Yes. So we've had to grow our R&D organization over the last few years, adding a ton of data science, machine learning, machine learning engineers, not just machine -- data scientists, different things. It's always like data scientists, but you actually also need the ML engineers and the AI engineers to implement these -- the algorithms that we create. All of our algorithms are created in house. They're all built by us. We have many, many patents against a lot of the technology we've created. We've had to definitely build up, over the last few years, a new expertise and new parts of the organization. Those are net new parts of the organization. Now when you look at the overall R&D organization, ever -- everything is -- we've built a platform. So we'll make enhancements to Knowledge Graph. We are enhancing the Answers experience, the Answers product. We are making enhancements to the core platform. We are making enhancements to Answers. Same thing with -- when we make enhancements to Pages and things of that nature. So when we talk about sort of the split of Answers versus Listings, it's not really thought of as Answers versus Listings. It's really that we have a generic platform with many different platform services, and those platform services can be pieced together to create a site search experience, to create SE-optimized pages, to potentially send data to endpoints like Google My Business, like Amazon, right, like our publisher network. And so that's kind of how we split it if we kind of look at it. It's not exactly like the sort of like here's the Answers team and here's the Listings team, and they're sort of looking at each other, and they're not -- it's not adversarial at all. It's actually that it's by kind of moving away from having a set of point products or even a multiple set of point products and moving towards a platform, we've been able to, in essence, sort of extend and broaden the R&D capabilities of the organization and ultimately creating an incredible amount of leverage, which I don't think you guys have even seen yet, which we're excited to sort of show you as we can now create solutions on top of the platform as opposed to having to require our engineers to make changes to our products. So this will be really cool.

And maybe, Steve, if you could just comment around the long-term model part of the question.

Yes. We had in the model around 15% R&D. Keep in mind, there's capitalization of new software in there. But as Marc said, as we need to bring on new people, we're going to be able to do that just like Dave can bring on quota-bearing headcount. But there's also this efficiency where large R&D teams don't necessarily give you the most brilliant solutions. And our team to date has been wonderful at that. And I think we're not hindered by being able to spend money in R&D, but we'll spend it wisely, just like we're going to hire wisely into quota-carrying sales reps.

Great. Thanks, Steve. And with that, we'll conclude our Q&A session, and we will conclude our Investor Day. Thanks so much for coming. Thanks for your time and attention. Have a great day.

Thank you.