Salesforce, Inc. (CRM) Earnings Call Transcript & Summary

Good afternoon, and welcome, everyone. Ali and I had the absolute pleasure of presenting our connections in Chicago a few months ago. And we are presenting this session via cloud Email marketing, not big, but it was quite a hit, I would say, it was a full house standing room only, and we have reached to us that themselves. Now also at Salesforce Connections, we announced some pretty exciting innovations. So I'm going to keep seeing what were some of those announcements, we'll take a look at a bit of a similar. Then I'm going to hand it to Eliot [indiscernible]. I am Linda Nguyen. I'm a lead solution engineer for marketing cloud based in Sidney , Australia,And if you are me as well, it is very warm today. bit kind of worrying.

Thanks, Linda. Yes. I am with Cloud Capital, who are Canadian-based Salesforce partner located in Halifax in Canada. But I'm actually based in Orange in New South Wales. I am a Salesforce MVP and for those of you unaware of what an MVP is, it's an award given by Salesforce to recognize experts and their contribution to the community. I've received that for the past 6 consecutive years and recently been inducted to the hall of fame. It's a privilege to receive that recognition and award. I've been using Marketing Cloud now for over a decade and written books on Marketing Cloud, including the [indiscernible] Guide, which is my latest book.

Well, there's some very impressive accolades, a publisher was [ in my control ] a little bit more. Looking around me, I can tell you that I was also SEO of the year once Salesforce, what else can I assure you, got a couple of medals from some Salesforce Olympic Weightlifting that I competed in, yes always a thought of intro, a bit more.

I did my own [indiscernible] medals actually. So this is I think you need to be a weightlifter to lift them up. So this is every single Dreamforce connections, event world to all the different Salesforce events, I've been to it for a past decade. So it kind of plugs up my wardrobe.

That is incredible Eliot, very, very impressive. I would like to see you wearing them all around your neck, but I worry that's a health and safety issue. So maybe not. Thank you very much for joining us today, Eliot. Now it wouldn't be a Salesforce presentation without this legal slide. So before we get into what is coming in the future, this is a forward-looking statement and it's asking you to make all of your purchasing choices based on what's available in the platform today. Now we are going to be talking about what is coming. So if you are looking at buying Marketing Cloud, actually buy base is available in the platform today, that's readily available, which is everything that Eliot is going to be talking about. Now what did we announced at Connection, we announced marketing GPT, and it is the first GPT powered marketing platform that is built on the world for it's trusted CRM. We are reimagining Marketing Cloud with generative AI capabilities. It's going to transform highly interactive with in marketing setups. It's going to allow you to fuel creativity with data-driven answers. So with that, you're going to be able to use our trusted customer data to fund things like campaign strategy, audiences, content. We're going to be able to unlock efficiency with GPT-enabled campaigns, and this is going to save you how countless hours on every step of the campaign life cycle. You're also going to be able to redefine the limits of personalization at scale with the combination of GPT created content creation and predictive AI recommendations, pretty powerful [indiscernible]. Now let's take a close look at what is coming because it's going to be coming a little bit closer than you realize. Segment creation, now this is currently in pilot as of last month. And this is going to help you build data cloud segment, faster than ever, just using natural language products. That's where you go. Next up, we have Email content creation. Now this is going to be a pilot in October, and this is going to allow you to ask Einstein to create personalized email subject lines and body copy variants and you can either choose from them or improve upon them. And lastly, we have the type-based partnership. So this is using generative AI, so you can create and access contextual assets that you can use for your multichannel campaigns. And we're going to see all the business action in [indiscernible]. Now before we get into that demo, I want to spend a minute just talking about how Einstein GPT grounding was because this is what makes the difference. Now it all starts with a prompt and then you ground with your trusted customer data in your CRM. So in this step, it brings an important data that's stored data cloud, but it also ensures data security and privacy. Now when the large language models run early enough information to make a decision is shared, and then it is chaso. It's not stored in external systems and your data is not mixed with public data. I'm going to repeat that because it's very important. Your data is not stored in external systems and your data is not mixed with public data. Then finally, the generated content, it's going to be more accurate and more personalized because of this grounding. Now I'll show an example and we will take a look at how Einstein GPT grounding makes the difference. I'm going to start off with this prompt, that's asking help me regarding catch Email to advertise connections to 2023. Now let's take a look at our results with no grounding and the result is GSM Connection's 2023 the essential Salesforce event for marketers is just around the corner. Seats are going fast, but there's still time to register. If you have attended before, we have reserved you a ticket. Yes, not bad. Now let's take a look at the result grounded with CRM data. Give them. I am thrilled to invite you back to Connections 2023. Seats are going fast, but as a valued Marketing Cloud customer we've saved 4 tickets for you and your colleagues at [indiscernible]. Now both of these Emails generated by GPT. But you can see that one on the right, the version that's been grounded with CRM data, it has details like the fact that the customer has been to Connections before that they are a long-time marketing cloud customer that's all been added. Now to reiterate, we are not going to be using new customer data to train our AI models. We have a zero retention policy so that your data is safe and secure. So that's a little bit on how Einstein GPT grounding works and how it makes the results more accurate and personalized. So let's see marketing GPT in action. I need to practice. This is a forward-looking visionary demo. It's going to go beyond what's been announced. But the idea is to get really excited about marketing GPT, so let's get into it. Now as we all know, market is on the hook to deliver greater growth with greater efficiency. Now fortunately, for our market of the future Oven at Northern Trail Outfitters, he can just ask Einstein GPT. Now she can ask to recommend what products she should promote so that she can maximize revenue from available inventory. Einstein GPT we use Northern Trail Outfitters trusted customer data to make a recommendation, and it looks like it's footwear . Now she can also ask Einstein GPT to help her create a campaign so that she can turn this insight into action that launch that campaign assistance. Now first thing is first. Behind any good marketing campaign is a data-driven [indiscernible]. And that is exactly what Einstein GPT has created here with using the conversation from Slack as that starting point for the prompt. Now it is all interactive. It's fully editable as you can see here. Now after the briefest generators, she sees that Einstein GPT has automatically drafted a full campaign based on the information that has been in that brief. Now what makes us unique is that the campaign is grounded with a trusted customer data in data cloud. So it means that the answer that she is getting, they're accurate. The data is secure and no hallucinations for Einstein. All right. Let's dig into this, and we'll start off with the audience segments that are created. Now after reviewing the salesforce , for this segment, our marketer decides that she wants a silo broader audience. So she asked for additional activities to be included, things like camping, hiking, biking and Einstein GPT translates to natural language and through the appropriate segment attributes. How you [indiscernible] now marketers also tell us that one of the biggest challenges is actually creating content for personalization. Thankfully, Einstein GPT is one step ahead, has already created a draft Email for a view. Now let's take a look at the content. Not bad, it's on the right tracks, but among the things, the copy could be a little bit more exciting, that's better, I like it, a bit more sizzle, but maybe a little bit too long, she asks Einstein GPT to make it shorter. Perfect. Now that she is happy with the copy, she can now turn her attention to the images, which could probably be a little bit more exciting than simply is on the white backgrounds. Now this is where our new partnership with Typeface comes in. Now she's using the same chat that she had with Einstein GPT, and she can ask to add a nature background to that product recommendation block. Now with Typeface, Typeface are going to generate multiple image recommendations based on Northern Trail Outfitters brand content and guidelines. Our marketer selects the image that she likes best and she saved the Email. How much better does that look? Now we are not done yet. Einstein GPT has taken care of the activation and created a cross-channel journey, and it's all based on the campaign brief that she started with like things like converting loyalty programs to add it in there. So now it's in the journey. Well, She thinks that the journey looks good, so she hit save and activate. Now with the marketer in the loop at every single step, Einstein GPT has helped to create a personalized campaign faster than ever. How good does that work? Now that is a taste of what is to come with marketing GPT. And as I mentioned, some of these features, they currently in pilot or they're going to be piloting soon so watch this phase. Now let's not forget the reason why we're here, you can actually already start to use AI and marketing cloud right now or probably not right now, you're in the webinar. Now we have multi machine learning models requiring a variety of tools in the platform. It spans things from natural language processing to propensity scoring, better yet, marketers are going to be able to leverage these tools using clicks, not code. And so to help you with things like segmentation, this does help you with continent and copy, even insights. So, I'm going to handle over to Eliot from Cloud Capital. He is going to deep dive into how you can unlock value with AI-powered Email marketing. Eliot, over to you.

Thanks, Linda. Now before we can understand how AI can help you in your Email marketing campaigns, it's helpful to ask some basic questions. And to do just that, I'm going to be using the Five Ws, which is a popular information-gathering framework that's been used for centuries. In today's session, we'll look at answering the following questions. What text and image content will keep your subscribers engaged? Who should you be sending Emails to? And what are the different treats and personas of your audience? Next, why should you optimize yourselves? So instead of blasting all your subscribers with Email like a blow fight canon, how many Email should you be sending to each subscriber? Then when is the optimal time to send an Email to each subscriber to maximize your open rates and engagement? And finally, where can you go to find out when your Email engagement rates become an anomaly and require intervention? Well, the good news is that AI can help you answer all of these questions. And in fact, most marketing organizations are already aware of the benefits of AI as in a recent survey last year, 68% of marketers say that they have a fully defined AI strategy, which is up from 60% in 2021 and 57% in 2020. Well, let's now unpack each of these questions in turn. Let's wait for the slide to load. Firstly, how can AI help to determine what content to send? Well, image libraries can quickly become disorganized and unwieldy and manual image tagging is a very time-consuming and subjective process. And it can be challenging to locate relevant images from an ocean of assets, which can be likened to finding a polar bear in the snow storm. Now this typically leads to content fatigue where marketers revert to using the same image regularly. But using the power of machine learning, Einstein content tagging, automate image tagging by identifying objects and themes in your image assets, building a searchable content library. And this increases the efficiency of your content search so you can most easily find the most relevant images for your message. Einstein content tagging adds an update content tags and meta-data automatically, every day. In addition to the tags that you've added yourself, so you can quickly search, identify and organize your different assets. And similar to AB tests for Email, Einstein content testing lets you quickly and easily test how your image assets are performing. So instead of analyzing which links customers clicked on in which Email, you can use this feature to test up to 32 images and determine which images work best. You can either create manual ABN tests or automated tests, where winning images can be dynamically selected based on the highest click to open rate. And for more complex targeting, Einstein content selection includes fatigue walls and targeted content based on user preferences. This AI feature solves the problem of not having enough data or programming expertise as image selection is driven from your content pool. We've built in ABN testing that optimizes in real time. So Einstein constantly picks the best image assets more quickly. Now when we begin sending, Einstein chooses the best image asset for each customer from the assets and the customer information that you supply. Each Email is customized for the customer that is sent to using the assets you provided. Then as Emails are opened using real-time analysis, Einstein finds the assets that prompt customers to click on a link in the Email and select that asset more frequently over future sense. And this feature is showing impressive results across Marketing Cloud customers. In one case study, an Email that used Einstein content selection resulted in an 18% higher return on investment than Emails that didn't use it and in addition to determining what content to send, it's equally important to determine what subject line to use. If arguably, the biggest challenge in Email marketing is getting subscribers to open Emails. You see the subject of an Email is much like the cover of the book or the headline of an article, it needs to be compelling. Otherwise, you'll carefully personalize irrelevant content where it will head straight to the trash. And Einstein copy insights drives Email open rates through optimizing subject lines and enables you to understand the words, the phrases, and tone they get subscribers to open, while also comparing the impact of phases, tone and punctuation on subject line engagement over time. You can even test how your message a subject line is likely to influence messaging engagements performance, the for it sent and compare multiple subject lines. Next, how can AI help to determine who to send to? Well, Einstein engagement scoring is a predictive segmentation tool for Marketing Cloud that predicts how your audience is going to engage with your Email marketing campaigns and programs. Einstein scores every individual subscriber based on their likelihood to open, click, stay subscribed and convert. And once you can predict what a subscriber will do next, then you can gain audience insights by understanding the factors that predict engagement, identify trends and measure audience health. So you can see how you're engaging with your audience and how they are engaging with you and it also creates individually addressable segments, which are organized into targeted personas. You can then take action on this segmentation data using the Einstein Split Activity in Journey Builder to put customers on the right path based on their persona and predictive behavior, optimizing the journey for each individual contract. Something has happened with the slides there. So let's jump back and looks like it skipped ahead. Great, It's got a mind of it's own today. So here we go. Next, why should you use AI to optimize sending. Well, with customers receiving more message to their inbox than ever before, marketers need a better way to manage their Email engagement strategy. And Einstein engagement frequency provides an insight of how many Emails you should be selling to your subscribers. And similar to frequency capping, this feature allows marketers to understand the optimal number of sends and at what point subscribers are becoming oversaturated or where they're being under-messaged. Essentially, it identifies the sweet spot of your email to send ratio, allowing you to send the right number of messages to optimize engagement while also reducing the chance of email fatigue. Einstein engagement frequency helps marketers reduce unsubscribers by finding oversaturated subscribers and also tap into additional sales opportunities by finding undersaturated ones. And you can then action this engagement with [indiscernible] frequency split activity, which uses real-time engagement analysis to automatically segment contacts in running journey and send them down an appropriate path based on the saturation level and when is the best time and day to send an Email, deciding when to send an Email can be a highly subjective process, and there are many different factors that influence when someone is most likely to open an Email. Well, Einstein sends time optimization uses machine learning to predict when a subscriber is most likely to engage with your message by analyzing the optimal engagement time for each subscriber to maximize open rates and engagement. Whether you're sending Emails within a journey or scheduling batch sends in the automation studio, Einstein Send Time Optimization will assign the best hour to send to each subscriber to help place an Email above the fold in the inbox to capture attention. In summary, it removes the guess work that is determining the optimum send time. And the Einstein Send Time Optimization dashboard that we see here provides analytics to predict future message engagement. And finally, once all Emails have been sent, where can you go to look to monitor KPIs? And what should you look for? Now Marketing Cloud already tracks Email engagement metrics at a very granular level. But when does an open or unsubscriber become an anomaly that's statistically relevant enough to require intervention. Well, Einstein Messaging Insights provides anomaly detection for email marketers through a notification center where marketers can review irregularities in KPIs. So if there's a sudden drop or gain in the Email engagement, for example, opens clicks or unsubscribe rates, then Einstein Messaging Insights will service these insights on their marketing cloud dashboard. Insights are updated daily, so these automated notifications aren't reliant on a marketer reviewing email tracking data or scheduled reports. In short, Einstein Messaging Insights enables marketers to stay on top of Email performance changes and use provided insights to optimize journeys and absence. Let's discuss one possible scenario. Perhaps in a banding cost journey has been running for some time, but a recent update of dark mode for iOS has impacted click-through rates on these devices as the design no longer works as well in dark mode. Well, in this scenario, Einstein Messaging Insights with extended notification directly to the notification center for that past 24-hour period alerting the marketer that there's been a sudden drop in click activity, so they can identify the cause, intervene and apply the updates to ensure the Email performance stays level or increases. In summary, the benefits of AI are clear to marketers when you can filter that AI uses data to make informed decisions. AI can perform specific cognitive tasks as well as or better than humans, unlocking deep insights, resulting in increased opening click rates, while also would like to insights and patterns from large customers the personalized experiences. In fact, in a recent survey, [indiscernible] customers expect us to always be personalized. And AI unlocks hyperpersonalization use cases, so you can use behavior and real-time data to create high textual communication, changing your messaging and conversation from what was to what's next as you can now anticipate your customers' behavior. And as Linda mentioned earlier, the best news is if you already have marketing cloud, then you can start using these AI features today or perhaps tomorrow and unlock value from your marketing campaigns. So I have to learn more about the Einstein features that I've discussed today and for other marketing cloud tips, I'm head over to my video blog at mc.chat. We all see links to more of my content, including Cloud Kettle's YouTube channel. Back to you, Linda.

Thank you very much for that, Eliot. Thank you so much for sharing your wisdom. Now I'm going to share with you a bit of our road map, you can start to add AI to your journey. Now when we're looking at the first 2 features, the first step, the core phase, simply just turn it on, turn it on, let the insights flow in the end. So let me give you ideas on the types of and carrying the messages sents in the space . You can also automatically create segments that you would have had to manually do in the past. Now when it comes to the work phase, this is all about adding segment splits to your journeys or adding content to your ongoing campaigns. Now the run phase becomes a little bit more advanced, but definitely achievable. In this stage, perhaps you want to test the number of messages that you're selling to segments with engagement frequency or if you're testing content selection, it's not sharing more information with the model so that it makes the recommendations more tailored in the experience ritual. Now I'm going to leave this here for a second. Then take a screen shot if you want to might be it is how to get started, but AI is here to help and it's something you can start doing right now or tomorrow. Now before we finish up, we're working on today, I want to thank you to offering Robin. He's been answering your questions throughout the webinar. And we have had a few come in, so I'm going to take a look and share some of these.

So let me take a low but start off with some of the ones on marketing GPT. So one of them is, when is marketing GPT going to be available? The dates for the general availability of marketing GPT, not confirm just yet, but I did mention that with things like segment creation and data cloud, that's already in pilot, Email content creation that is parting in October. So no doubt, we'll probably get some updates on the time lines of Dreamforce, which is just around the corner. So I would say watch this Phase. All right. Well, let me see what else I can find here. This is good one. Does Einstein AI and marketing GPT come with all additions of Marketing Cloud and how much does it cost? Well, marking GPT be the full details are the yet to come. But with the AI features the Eliot spoke to , things like Send Time Optimization, engagement scoring, all that stuff. If you're on corporate or at a high positions of Marketing Cloud, that's already included. If you are on professional edition and you want the AI features, then you can just purchase it as an add-on or depending on your needs, you could upgrade your edition, depending what you need like if you upgrade, then you can unlock things like Journey Builder, what of our messaging things in SMS and push notifications. But I would say, have a chat to your account executives about your needs and they'll point in that direction. All right. We've got a few more questions that I want to share. And Eliot, I'm going to send some your way. There is one here that is asking. I don't see all Einstein features enabled on my account. How can I provision them?

Yes, that's a great question. So look, several of these Einstein features for them, including Einstein Engagement Frequency, Einstein Engagement scoring, Send Time Optimization and actually also Einstein Content Tagging aren't enabled by default. But the good news is they're easy to activate. You can just go to Marketing Cloud setup and there's a button to go and activate them there. On the side note, note that for definitely Einstein Engagement Frequency and Einstein Send Time Optimization, you need to wait at least 72 hours after activation before you can start using these features.

Okay. Noted. Thank you very much. Let's point you one more question and is on Einstein Engagement Frequency enabled for all business units and send classifications?

Yes. Okay. That's another great question. For this Einstein feature, specifically, no, it's not. You need to activate the feature in each required business unit. And to answer the second half of that question, is it available for send classification. So only intentionally, only commercial sends are analyzed, not transactional, as transactional sends typically have a much higher open rate and skew the model.

That makes perfect sense. Everyday is a schoolday. There you go. Always learning something newer. All right. Let's leave it there for the questions. And before we finish up today, check out the related content tile in the webinar platform. We've thought a whole bunch of resources in the full year. So there's things like you can learn more about Marketing GPT by the Trailhead link. We also shared with you 50 best practices for Email marketers. That's practically a new best practice every single week for an entire year. So that will better keep you busy. You can also check out the top 10 KPIs for Email marketers, spoiler alert, open rate is not one of them. And you can also check out Cloud Kettle's YouTube channel or Eliott's blog, so you can learn all about marketing cloud. Thank you again very much for joining us today. Thank you, Eliot, for sharing your wisdom. Thank you, Robin, commanding the Q&A.