Appian Corporation (APPN) Earnings Call Transcript & Summary

Hi. Thank you, everyone, for attending. I'm happy to have Appian with us today. We have their Head of Investor Relations, Scott Walker; and their CFO, Mark Lynch.

To start off, I think we'll go with just a high-level question. Maybe, Scott, if you want to answer this. For investors maybe new to the story, it would be great if you can introduce yourselves, provide an overview of Appian's history and maybe a brief explanation of what low-code is?

Sure. I can give the overview of the -- of what low-code is. So Appian is a low-code automation platform that allows companies to build mission-critical, complex applications up very quickly, 10x quicker than with traditional high-code manners. So that we allow customers -- we have global pharmaceutical companies that are building clinical trials on Appian. We have a lot of -- a variety of different complex use cases, allow companies to connect to various legacy systems and data to service orchestration layers and being able to really help them solve complex problems.

And to kind of talk about the history, I've been with the company for 13 years. I've known 3 of the 4 founders from a previous life at another software company. The company was founded at 1999, and it was bootstrapped into 2008 where we just took on $10 million of primary equity until the IPO. So we literally ran the business on the backs of our customers. We started off kind of a solutions consulting organization. And we got into BPM, and we actually -- one of the first things we've built was the Army Knowledge Online portal that basically serviced 3.5 million military individuals and their families. And it basically allowed them to communicate with each other and find out where they were being transferred and when they were getting deployed and all that stuff. And so we ended up learning how to handle some of the toughest and most complex things with software. So we had to deal with scale. We had to deal with concurrency. We had to deal with security because this is like federal government, federal government DOD. And so we -- instead of starting from the low end and building simple workflows, we actually started at the high end and built some very complex applications. And we took what we learned with AKO and basically transferred that over into the platform. So a lot of the foundation of the platform came from the early days of the company's history.

Helpful. And then when you say you're building complex applications, is that -- can you just walk me through like how the actual developers are using Appian? Is there like a dashboard they're using at their desk? And I guess what tools do they have in the Appian application or platform that allows them to make these applications super quick?

So there's a lot of different things. We have prebuilt user interface. So if you think about one of the more complex things to do in software is to build the user interface and make it beautiful. So it's like 98% modern-looking. It's prebuilt right out of the box. We have a ton of prebuilt wizards. We have a lot of integration. We have low -- zero-code integrations with dozens of systems like Salesforce, SAP, Oracle, et cetera, NetSuite. So people can just plug in immediately with those particular systems. The ability -- like, basically, the way you build applications with an Appian is like drawing a picture. So you don't go in and code. We don't allow you to go in the code. Appian is built on Java, right? But when you work it with it, you just literally drag and drop workflow pictures, and that's how you build out your applications. Every once in a while, the reason why it's low-code versus no code because every once in a while, you have to actually write some code to integrate with a complex -- like a system that -- where we don't have prebuilt user interfaces. We also have the ability to write an application once and have it be deployed natively on any device, any mobile device. So if you look at it, software engineers are hard to come by, and mobile software engineers are even harder. And so we take care of that problem for the CIO. The right lesson is to deploy natively on any device regardless of what you're going on. And we also -- we're zealots of security. So our system has all the security that you would expect. So we're PCI-compliant. We're HIPAA-compliant. We basically got IL4, which is the higher-level security within the Department of Defense and federal government. So we have all the securities certifications that you'd need if you wanted to basically deploy this either commercially or within the federal government.

I mean that's a helpful overview. And then maybe just broadly, what stage of adoption do you think low-code is? I know you guys have been around for a while, a couple of decades. But it seems like over the last, call it, 12, 18 months, we've really seen an inflection point in terms of just number of people talking about it and even in your business. So interested to hear your view on that.

Yes, we were the first pure low-code vendor to go public. When we IPO-ed in 2017, our CEO likes to say that he raised the curtain on low-code. I think you're exactly right, our theme of our earnings call a couple of weeks ago was that low-code has gone mainstream. And in fact, the analyst at Forrester estimates that by the end of 2021, this year, that 75% of companies will be using a low-code platform. And I think a couple of things have happened, particularly in 2020, to help drive this is, I think that, with some of the impact of the global pandemic, is that companies understand the importance of digital transformation and the importance of software and needing to automate manual processes and then also change. Companies went from working one way on a Tuesday to a whole different way on a Thursday. And having a -- being nimble and being able to react to change became very important. And with a low-code platform like Appian, we're able to be deployed quickly and, even more importantly, be changed very quickly. And I think that's really a great tool for companies to be able to react to change.

Perfect. And then I think you guys have also kind of moved into the RPA space, which is also, I guess, very hot with an acquisition about a year ago. Now you have your own kind of Appian RPA, but you also integrated with much of the more well-known RPA platforms like Blue Prism and Automation Anywhere. So I guess, how much we think of Appian as maybe an RPA product? Or is it just like one feature of it? Maybe what are your broader strategies within RPA going forward?

I would think of RPA as a feature within the Appian platform. We're not selling on a stand-alone basis. But you're right, we purchased Novayre early last year. Their product, Jidoka, was the #1-rated RPA product on Gartner Peer Insight reviews, and then we released Appian RPA during the middle of last year. And one thing that really differentiates Appian is we have complete automation capabilities. So that's being able to combine people, combine RPA bots, AI and various technologies all within a workflow. So we think of RPA as an important part of that automation story. And we still do have the partnerships with the big 3 RPA firms. And a big reason for that is most of these global enterprises, these large complex companies, have more than just one RPA vendor within them. So if we want to be able to work with these big 3, if they're already within the enterprise and the customer does not have an RPA, but within the use case needs RPA, then for convenience, we have RPA available to them. For example, one of our long-time customers last quarter, they were -- they're, let's say, a Fortune 500 consumer products company, that they were -- they purchased Appian to do product approval life cycle. And they had legacy systems that needed -- there was no API calls. So they purchased Appian RPA to be able to use to go and grab and fetch the data to be able to pull into the workflow. So it's been a successful offering, and we're very happy with our Appian RPA product.

And I guess maybe just when you think of, like, developers that are using your products, I guess, what is their view of Appian in terms of how does that affect their day-to-day job from not using Appian into now using Appian? Does it make it much easier? Is it kind of addictive? I guess any comments on that?

It makes them far more productive. I mean we -- with Appian, you can build an application 10x faster than with traditional Java. And there's customers that -- and these companies have backlogs of projects that they're working on, that with Appian, these technical -- these IT professionals are able to now solve more problems and build more applications than they would just under a -- just using existing high-code measures.

Okay. Maybe switching gears a bit to the partnerships. You have a very established partner network with KPMG, Accenture, Deloitte, PwC. Could you just talk a bit about your partner strategy and how they're building their own services around your product? And maybe where that partnership strategy has gone from like a year ago to where you think it's going to be in 3 or 4 years?

Yes. I'd say we're still in the early innings with the partner ecosystem. Having said that, 70% of our new logos we got last year were partner-influenced. So that was great. We'd love to see the traction we're getting with the partners. The business relationship is that the partner will bring us into a customer as a technology that they feel can solve a particular problem that their customer is looking for. The partner is trusted with that customer. They generally have that C-level relationship, and they get the services business. So we go in, we sell the software to the customer, then the partner actually builds the application or applications, depending on what the customer is looking for, on Appian on behalf of the customers and they get that revenue. And that was one of the reasons, if you looked at our services revenue over the past 2 years, it's kind of flatlined, if not declined a little bit. And that's predominantly due to the fact that partners are involved in more and more of these deals. And we also had a little bit of COVID headwinds during 2020 as well.

And I guess for the partners, it seems like they're building a practice solely around Appian. Is it more so that? Or are they more so lumping Appian into maybe a digital transformation practice? Or is it like an Appian-owned practice?

It's both, right? So one partner has a digital transformation practice. We are one of the technologies featured within that practice. We're the low-code provider if the customer is looking for that. But other partners with -- and any other partner within the partners, right, are literally building the practice, building an Appian practice within, for example, within KPMG. We have some international partners that are literally building practices, training their employees up on Appian, and then they're going out and basically getting deals, right? So they're going out to their customers, pitching Appian and closing the deals and we go in to sell the software, which is -- that's a great relationship. Partners are also more and more building applications on top of the platform and selling that application, selling that IP as well. So we would get the sale -- basically, they build -- for example, KPMG built a LIBOR application and they've been selling that to various customers. And the LIBOR application basically mirrors Appian and AI that allows them to read through all these contracts looking for LIBOR dependency language and then basically offering up a replacement language as the LIBOR obviously no longer exists. And so when they go in, they sell it, they get a certain amount of money for that IP. And we obviously get money for the software or the platform because you need both to operate it. So we're starting to see more and more of that happening within these partners. And the partners, I mentioned KPMG, they're our biggest partner, meaning that they've got more people trained, and they're bringing us more opportunities. Accenture is kind of the rising star within the ecosystem, and we also have good relationships with Deloitte and PwC as well.

And from the partner perspective, kind of developing that own solution, I would assume that's much higher margin for them than maybe more professional services-oriented, right? So this is kind of an incentive for them to go that route.

Exactly. Exactly. And we love it, right? If they -- and then what happens with these prebuilt solutions, it makes that first sale easier, right? The customer gets -- if you're trying to sell a horizontal platform, you're asking the customer to use their imagination on what we can do, right? That's a tough sale. But if they have a particular problem and you have a solution, prebuilt solution that can solve that problem, they'll basically get it. They'll buy it. But once we land, then we basically pitch the virtues of the horizontal platform, and that's how we expand.

Got it. And that, I think, is a good segue into pricing. I guess, how has pricing evolved over the years? I guess in my understanding that the customer has a choice to pay on a per-user basis or a per-application basis. I guess what causes one to sway in one direction over the other? And I guess what has been the general trend been kind of over the last couple of years? And I guess where do you expect it going to be?

So historically, we did per-user-based pricing. That's pretty much what a lot of the software vendors did, and we didn't want to come out with some maverick idea because we're a little unknown Appian. And so I would say now -- even now, about 50% of our contracts are per-user-based pricing. And the beauty of that is you can basically go in and build one application and they roll it out to 100 people. You build another application in a different department, you roll it out to another 100 people. You get the licenses for those. The thing the customer likes is they can build out applications and then roll additional applications out to the same users and I have to pay additional license fees. So our CEO, who's an economist by training, came up with a different idea about 2 years ago, and that was an app-specific license, flat apps is what we call it. And it's priced basically on a small, medium, large, extra-large like T-shirts. And really -- and the pricing depends on number of users, complexity of the application, the ROI, the derived ROI, et cetera. And all of that pricing is really done by a central group of people within Appian. And so we rolled that out about 2 years ago, and that's -- the vast majority of the deals that we've closed since then are application-specific licensing. The beauty of that is you sell an application, and you can -- you basically can have a rollout. You build one application, come up with another application. You can have the same users using each application. They have to pay the company -- the customer has to pay for the license for each one. So that's -- and what we do is we give the customers the opportunity. If they rather buy in a per-user-based, we'll do that as well. So we're not going to just do flat apps. And some customers like me, for example, when we were with flat apps, I was skeptical. I was thinking -- like from a budgeting perspective, I would say like we're going to deploy Appian to 1,000 people. I know what the price is. I'm going to budget it accordingly. I have no idea of what an app will cost me, and it's harder to budget. But it's taken off. I think the customers has reduced the friction in the sales cycle, and it's taken off. So...

It's been then -- how does that kind of -- if you're comparing the 2 different pricing models from an upsell or a contract expansion point of view, is there any similarities or differences between what you might expect from customer A doing a per user and customer B doing a, I guess, an application model?

It's generally -- they're similar, right? So a lot of the expansion occurs when you build additional applications that are either deployed to additional users or those applications are just licensed by the customer. So that's -- our expansion is the exact same because that's where you get the expansion as the additional applications. And it's either licensed on a per-user basis or it's licensed on an app-specific basis.

Got you. And then maybe your customer base, could you maybe provide like an overview, just breaking down concentration by vertical? And then of that, I guess, what's SMB, what's enterprise and where you expect that to go?

Yes. I think it's important to remember that we're vertically agnostic. We've got great logos in a lot of different verticals, manufacturing, universities, health care. But our top 3 verticals are financial services, federal government and life sciences, which makes sense. You have 2 heavily regulated, heavy compliance-driven verticals and then the third is the regulator. So with Appian being able to be deployed quickly, to be changed quickly, we can help these companies and these verticals to stay up to date with changing rules and laws. So from an SMB, enterprise, we've got a very limited exposure to SMB. Our sales force targets the Global 2000. We primarily have -- most of our customers are going to be these large global enterprises.

Makes sense. And then your sales force, I guess, what's the type of investments are you making there? How should we think about headcount? What would new quota carriers, I guess, you might be planning on bringing on be focusing on in terms of vertical or customer size? And then how should we think about sales force productivity as well?

A lot of questions there. Let me maybe -- so basically, we talked about this on our earnings call. We're basically ramping up headcount in sales. During the year, during 2020, we saw sales cycles compressed by 30%, which was great. The sales force is getting more efficient. We're starting to feel a little bit like a flywheel, and it makes sense to invest aggressively, I think, in this space. So we're ramping up sales headcount. We've doubled our recruiting function to go get the sales rep sales count, and we're looking for quota-carrying reps both internationally, domestically, looking for sales engineers, looking for business development reps all across the board. Because it's -- we're actually -- there's a ton of opportunity out there. And we're focusing strictly on the Fortune 2000, predominantly on the Fortune 2000 or large businesses. We're vertically -- basically, we have -- we're vertically aligned in 3 different verticals: federal, pharmaceuticals/health care and financial services. Those are 3 larger verticals. So our sales force is basically -- we vertically align them to those 3 verticals. Some verticals that look promising would be energy, education, manufacturing, but it's not to a point where we need to actually make the investments to go further there. But -- so basically, the headcount is coming in. They're going to be focused on large businesses that are underrepresented right now. And we also have a new Chief Marketing Officer, and we'd be giving her a pretty sizable budget from a digital marketing perspective to go out there and really build brand awareness. I still think that Appian's brand isn't well known. Our customers love us. But when you think of low-code, you're starting to think of Appian, but I think we need to be top of mind.

And is that the -- I guess, with the 14-day free trials also helping to do try to get brand awareness out there?

Yes, the 14-day free trials are working great. 53% of the new logos that we got in 2020 went through the free trials first. And in fact, one of the -- an interesting story, we talked about RPA a little bit earlier, but one of the -- one of Canada's largest cellphone and telecom providers basically went in and built 30 RPA bot processes in Appian free trial before they actually purchased Appian. And what they ended up doing is they ended up replacing 40-plus Blue Prism bots with the Appian RPA. And they selected us basically for lower total cost of ownership, the fact that we integrated with the overall platform, the image recognition quality and then the fact that we integrated with Linux and SAP support as well. So that was an interesting win story there as related to RPA. But only 34% of the net new logos went through their free trials in 2019, so it's definitely gaining traction.

Is it maybe becoming a bit more self-service-y as well? Is that a decent way to think about it?

I mean it's a lot better than it used to be. Like I remember going in a couple of years ago and playing around us, I mean, this is embarrassing. Hopefully, nobody looks at this. So they've made it really good. It's really -- and the fact that you can go in there and actually build stuff with it is really good. Appian still, it's not like a citizen -- it's not for citizen developers, right? You've got to be a technical, IT developer kind of person to go in there and play around with it. But if you have that technical expertise, you can literally build stuff. There's other stories out there that customer -- like a 30-plus-year-old person went in there and built out the applications that they ultimately wanted to have within the company in the free trial, showed it to his boss and they ended up buying Appian because of that. So it's a useful experience. It's not like you download and you immediately start using it, but it gives people a sense of the power and the ease of development.

No, that really makes sense. Maybe switching gears to the solutions business. I know this was a focal point, maybe not so much on the last call, but definitely the few before that. How should we think about the solutions business kind of going forward? It seems like you're really riding a line between the platform business and then the application company with the solution side. So I guess how should we think about this evolving? Is this more of the bigger growth drivers, I guess, driving that cloud subscription growth over the next 3 to 5 years?

Yes, I think solutions is an important growth driver. I think in the short term, what it's really doing is helping to reduce sales cycles. It's also helping to bring us into customers that maybe they didn't -- or weren't comfortable making a platform purchase but are willing to make that point solution. And we did -- as you kind of called out, 2020 was a really positive year for solutions. We had COVID solutions, one being Workforce Safety, helping customers to safely return employees back to the workplace that brought us dozens of new logos. We have a suite of solutions for the federal government, acquisitions requirement, award management that's around the procurement process. For example, actually, the U.S. Air Force, almost everything they buy, everything but the complex missile systems are bought through Appian. So we're able to then kind of productize that and sold that to a number of new customers as well. So we'll always be a platform business. And the solutions are built on the platform, which we think is very important, that way when the platform is upgraded, the solutions are upgraded. In terms of building new applications or new solutions, you're able to use the reusable components in the solutions to build or even transition into a platform purchase. But I think that going forward, that will be an important part of the business.

And then I guess what influences you guys in terms of, we should make a solution focused on procurement or HR or legal? I guess is it hearing what your customers are saying? Yes, that would be great.

Yes, 2 main reasons. What are things we do really well? Like our first solution was institutional onboarding for the financial -- for financial services companies. They're onboarding complex customers, hedge funds, things like that nature, which we have built out a number of those applications. So it's what we do well, mixed with what are pain points within these industries. What are the problems that customers are having, and we'll kind of marry those together, and that would make a good solution.

Do you think going forward, you are going to be making more solutions on yourself? Or do you think it's going to be more partner-led from a solution perspective?

It's probably going to be more partner-led. There's just more partners. PwC has solutions built. Mark had mentioned the LIBOR solution for KPMG. There's just more of them building solutions, but we'll continue to build our own solutions. And kind of one way of how we build solutions is, it's almost as if we'll have kind of a family of solutions, like we're seeing in the federal government. We released acquisition requirements management and award management. It probably will work similar in the financial services. Institutional onboarding can then lead to a KYC solution to an anti-money laundering solution, and they can kind of click together and be these family of solutions.

Perfect. And then maybe just touching on the, I guess, the market and the competitive landscape. It seems like Appian is really in a ton of different sub-verticals of a lot of big markets. How should we think about the market size? How fast is that growing? And how can we kind of tie that into your cloud subscription revenue growth rate, which was phenomenal at 40% this year? But yes, I guess, maybe anything on that?

We look at the market size a few different ways. So we take our core software categories, low-code, intelligent process automation, our RPA, these areas where we're a leader. And if we add up the TAM for each of those markets, it comes to $70 billion. You can look at the custom software development market, which, at the end of the day, is what we're doing, and that's $230 billion. So there's a really big runway out there. There's a lot of space. And I think it's important to keep in mind that where we sit at the high end of the market is a good place for -- while there's competitors there, most of these other low-code vendors are more at the low end. They're not building these complex applications, which I think gives us a little bit of some protection at the top.

And how difficult for these other vendors would it be for them to kind of move up in complexity to become bigger competitors? Is it a matter of just integrating? Is it a user base issue? Is it a trust issue with customers?

I think it's a couple of things. I mean some of the companies that have come in at the lower end, the massive cloud providers, I mean they're great companies with tons of resources. If they want to spend the investment on the technology, on security, on the scalability to try and move up over time, they can do that. But I don't think they have to do that to still be successful. And it's a very fragmented market. There's not going to be one winner take all. I think there can be multiple winners and multiple winners within different spaces.

And when you...

And we've been at this, like I said before, we've been at this for over 20 years. We came up through BPM. So the workflow -- the key differentiator of us is the workflow. Nobody else has the ability to basically seamlessly integrate people, bots, AI and other technologies that we can and build complex applications. And to Scott's point, there's tons of opportunity out there to build simple little lightweight workflow within departments. The way we look at it is, if a customer wants to build out applications a CIO needs, if you look at their backlog, they've got a 3- or 4-year backlog of stuff they need to do, right? And they don't have enough software developers to do it all. And so we basically augment that development for the CIO. And a lot of our customers now recognize the fact that they can build a lot of these projects that are backlog in Appian. They can actually standardize a lot of their software development in Appian. And the fact that we can do mission-critical stuff and almost anybody else out there can't, it allows them to standardize. So they can build that mission-critical couple of applications they got, but they can also build the simple lightweight workflow that they could build in like a power app or something like that. So it gives them the ability to standardize it. The fact that we have the security perimeter, you can have all of your applications behind one security perimeter. You can be in cloud or on-prem, you have the ability to toggle back and forth. So we make it -- trying to make the CIO's life as easy as possible so they can focus on getting these applications built for the business users.

If you had to quantify maybe your lead from a platform competitive position, I guess, how many years would you say it would take for somebody to kind of catch up to you? Or is that not a good way to think about it?

Yes, I don't think that's a good way to think about it. Because as they are trying to catch up to us, we're making it easier and easier and easier to build applications, right? Our sole focus is low-code application development. That's all we do. Whereas all of our other competitors focus on a whole bunch of things, right? This is just one thing that we're looking at. This is our sole focus. The goal, candidly, is every 2 years, our CEO has got a goal that every 2 years, I want to be able to build an application twice as fast, right, or in half the time. And so if we can actually deliver on that, like Scott mentioned that right now, you can build an application over 10x as fast versus Java, 2 years from now, if it's over 20x, it's over 40x. There's no one that's going to be able to catch up to that.

Yes. No, that's a good point. And your customer retention is 99%. Is that similar with other low-code platform that's out there? Is that kind of unique to you guys? Is it very difficult to rip and replace maybe competitor solutions or -- and yourself as well? How should we think about that rate going forward? Because obviously, you're powering a lot of these mission-critical applications. And once they're built, they're not going to get rid of it. So it seems like your platform is extremely sticky. I was just curious if it's also as sticky for other competitors as well?

I have no idea what other competitors' retention rates are. The fact that some of whom don't talk about their retention rates may tell you what's going on there. Over the past 3 years, our gross renewal rate was 98%, right? So it's not just a phenomenon during 2020. And the fact that you're going through a pandemic and we have a gross renewal rate of 99% says a lot about the value that we're giving our customers because I know as a CFO going to the pandemic, I look for everything possible to cut, right? If I don't need it, I'm going to cut it because I don't know what this thing -- I don't know what this COVID thing is going to do to our business. And thank goodness, nobody cut us, right? We basically renewed throughout the entire year.

And even in terms of the rip and replace, one of our largest customers is a bank in the South that they had been a big customer of one of our competitors. They grew -- we were a little tired of them and ended up then having an RFP for a new application. We won that business, and they've built all future applications on Appian. They didn't have to rip out the competitor. These large global -- these Fortune 500 enterprises have a lot of complexity, a lot of different problems, but there's room for even multiple low-code platforms or vendors to sit within that enterprise.

Would you say that's pretty frequent where these Global 2000 companies are using you and, I guess, another vendor, and then there's maybe wallet share gains kind of as you guys progress?

I mean I think...

Oh, go ahead, Scott.

Go ahead -- you go ahead, Mark.

I don't think it's very -- it's not super common. It's generally one or the other. And the fact that a lot of times we're going in, it's a greenfield opportunity. So they haven't used low-code to build it out on says a lot. But we could be in situations where -- if you think about it, Microsoft is ubiquitous. They're everywhere, right? So you can have a department within a large enterprise where they're playing around power apps, and yet we're being deployed in another department building mission-critical applications. So that could very well be happening today.

I think the greenfield aspect is interesting because I think there's a lot of technology out there that enterprises are maybe just starting to realize that they could use, but they're not diving head first in the deep end. They're going to the shallow end and kind of learning how to swim per se. Would say that's pretty similar to what you're seeing, it's a slow start, and then once they realize the benefits, then they really ramp up their use and reliance on the platform?

We see that a lot. I mean that's how -- that's our expansion rate. The 119% is that expansion within the enterprise.

And that was a good uptick the last quarter. Is that something we should expect to kind of continue to trend up?

We've been saying since the IPO, we expect to be within 110% and 120%. It's going to bounce up and down a little bit. We do big deals. And so a big -- a few big deals can move it one way or another. So I would expect to stay within that range. If it goes below 110%, that will be concerning to us. If it goes above 120%, our CEO will be happy. So we'll see what happens.

You guys long said that you want to be a 30% subscription grower. That's roughly about where you're guiding to this year, which is a bit of a deceleration compared to what you guys did in 2020. As we look at that kind of growth target over the next 3 to 5 years, what do you guys need to do on your end to kind of hit that level from a maybe a partner capacity, a sales force capacity, customer expansion? I guess what's embedded in that target?

I mean it's all of those things, right? It's, I think, continued support by the partners, bringing us in additional opportunities, more new logos. We added 50%. We had a 50% increase in net new logos this year versus last year at 167. So more logos will help. Investments in the sales and marketing will help. And 30% plus, that's an elite growth rate, right? So -- and we're not -- we've been saying, to your point, since the IPO, we've been saying we're going to -- we see ourselves as a 30%-plus grower. And we've done it consistently. And so we'll hopefully be able to continue to do that.

And then, I guess, maybe on the services side, what's your ideal mix? Is that going to be something as partners do more, it's obviously further better from the cloud subscription revenues, the services will be impacted? If we're looking out, should we think this is going to be like a 80-20 mix?

Yes, I think it's going to continue to shift more towards software. Either services will continue to shrink like they have over the past couple of years or they will grow, but they'll grow at a much slower rate. And it's hard to predict, really. And embedded in our -- implicit in our guidance for the year was the thinking that services may actually decline in 2021, similar to what they did in 2020. But we don't know. We don't have great visibility on the services piece, and we don't apologize for services. Services, we got 37% margins last year, right? And we're talking about 30% margins this year. It's a profitable business. The fact that we're in there with the customers, that helps the expansion rates within those customers and net revenue retention. And we also learn the pros and cons of the platform. We can basically get that information back to the engineering group so they can improve the platform. So there's a lot of benefits to it. We don't apologize for the services.

Got you. And then maybe just one last question for me. Just on the growth versus profitability question, how should we expect that going forward? Obviously, the end of the year, the last couple of quarters, you've made some good strides on the profitability front. It seems like the gross margin expansion is really dropping through. Should we expect you to kind of reinvest that gross margin expansion, just driven by the mix shift into OpEx to drive growth, given the opportunity is so big? But how should we think about that dynamic?

Yes. We're going to -- and we gave a guide that showed that we're investing in sales and marketing and R&D. We're -- both of those are very important. To your point, we're -- there's a lot of large companies that are getting into the space. Low-code is becoming a thing, and continuing to improve the platform is paramount to us continuing to succeed. So we will continue to make those investments in the platform. I mean, if you think about it from a cash flow perspective, for the year, we only used $7.6 million, right? And actually, in Q4, we were profitable from a cash flow perspective, about $5.8 million, right? But we're going to not -- we're not going to have that drop to the bottom line in 2021. We're going to try to continue to invest. But the spend kind of gave it -- the spend will be similar to what it was in 2019, which was around $26 million. That's kind of what we think we'll do, which isn't that much. If you think about it, we have over $250 million in the balance sheet. So I think from a liquidity perspective, we're in good shape. But to not invest right now, I think, would be a disservice to our investors.

Understood. Thank you guys so much. I really appreciate you guys taking the time to speak with us today. I really look forward to speaking shortly. I don't know if you have any closing remarks, but I really appreciate it.

No, thanks for inviting us as we always enjoy this conference.

Thanks, Brett.

Very good. Thanks, guys.