AYX - Alteryx Inc. (AYX) Needham Virtual Infrastructure Data Analytics Software & Cloud Communications Conference (Transcript)
2023-11-15 18:20:10 ET
Alteryx, Inc. (AYX)
Needham Virtual Infrastructure, Data Analytics Software, & Cloud Communications Conference Call
November 15, 2023 13:30 ET
Company Participants
Kevin Rubin - Chief Financial Officer
Suresh Vittal - Chief Product Officer
Conference Call Participants
Mike Cikos - Needham
Presentation
Mike Cikos
Great. Thank you to everyone for joining us with Needham’s Tech Week. I’m Mike Cikos, and I’m pleased to announce that with us today, we have the management team from Alteryx, the CFO, Kevin Rubin; and the CPO, Suresh Vittal.
Just for some quick logistics before we jump into it, but this is a fireside format. I’ve drafted up some questions on my side, but would like to make this as interactive as possible. I want to make sure we’re maximizing your time here. So if the audience has any questions, first, there should be a chat box under interface. Feel free to lob in your questions in that format. If not, feel free to e-mail me at mcikos@needhamco.com. I will do my best to make sure that we’re getting to that while we have the team here.
With that out of the way, Kevin and Suresh, thank you very much for the time and for joining us today. We really do appreciate it.
Kevin Rubin
My pleasure. Thank you.
Suresh Vittal
Thank you for having us.
Question-and-Answer Session
Q - Mike Cikos
Absolutely. So quick question here just in the interest of an intro or an overview. I know there’s probably some newer folks here to the name, some others who are probably dusting it off for the first time in a while. So at just a very high level, can we start with maybe discussing Alteryx, its position in the market is providing as far as providing value to customers, who are primary incumbents you’re going up against and who do you compete with when customers are thinking through the competitive landscape?
Suresh Vittal
Yes. Thank you, Mike. I’ll probably get started. Alteryx, we serve as an orchestration and automation layer for analytics across an enterprise. As we all know, our customers are wrestling with the problems of data fragmentation and data governance, being able to apply analytics for the right use cases across every different functional area. And so we are very unique in what we do, not just in what we do, but how we do it and who we empower on the journey.
We’re unique in what we do because the unparalleled nature of the Alteryx analytics automation toolkit means an analyst, a data engineer, a data scientist, line of business user can do analytics of any shape size, form and advancement as they choose to. We’re unique in who we empower in that we kind of cut our teeth in really helping the business users get the most out of their data and their analytics. And over time, we kind of start to serve the data engineers and the data scientists and the IT teams as well. We know analytics is a team sport and that means being able to enable every part of the organization to do the most with the data they have to create the insights and most importantly, operationalize those insights. We do that in an easy-to-use self-serve platform for our customers. The other place where we’re unique is we have customers across the entire spectrum of deployment models. So we have customers who have a lot of their data on-premise and need Alteryx to run on-premise, and we support that.
We have customers who are on their journey to the cloud and completely embrace the cloud. And so we support that. And regardless of which cloud you’re in, you can get – you get to run Alteryx in the cloud, whether it’s Azure, AWS or GCP. And then more often than not, most of our customers are in some kind of a hybrid format, whether they’ve got data on-premise, they’ve got data in the cloud, they’ve got cloud applications they’re dealing with and they kind of need an analytics automation and orchestration piece across all of those environments, and we help them do that better than anybody else.
You asked a little about the competitive landscape. The – our biggest kind of competitive topic would be customers would have built a lot of stuff by writing code and they do – they want to replace it with packet software. Other competitors who show up in this space, certainly Tableau has a prep product and the analytics product in the space. Dataiku is one of the smaller emerging competitors in this category. And then there are other tools that kind of provide bespoke capabilities to functional areas, for example.
Mike Cikos
Got it. Got it. Thank you for that. And I know we’ve also been asking a number of our companies today around GenAI. It was super-hot earlier this year. I think it’s still front and center for folks, even though I think generally, people are taking a more considerate approach for the on-ramp of what that technology means, right? And so would love to get your sense, but with GenAI, how does that in any way change the market opportunity and impact Alteryx’s, I guess, ability to serve up that value on behalf of its customers? And we can go into some different facets from there, but would really like to kick it off with that.
Suresh Vittal
Yes. The way I think – I’d like to think about generative AI and certainly AI more broadly and generative AI more particularly, will – what it will do is it will lower the barriers to adoption for capabilities like Alteryx. It’s been our mission to empower analytics for all for the longest time. And what technologies like generative AI do is they really help us accelerate that mission. We think about GenAI, which kind of gives us the opportunity in really broadly in 2 distinct ways. One, certainly, it enhances our existing offering. So it makes it easier for us to do – help our customers do some of the work that they embark on already, faster to get data, faster to clean that data, faster to create analytical models. What could be better than to use English as your programming language, for example, and create these complex workflows just by dictating a set of instructions in English. And so we think that lowers the barriers to entry and really enhances our current offerings.
What it also – what generative AI will also do for us and has already started to show some signs of that is it enables us to address multiple personas. I kind of mentioned earlier that we very much view analytics as a team sport inside the enterprise. Data pipelines are complex. There’s data everywhere. There’s metadata that needs to be recognized and rationalized. You have to kind of create analytical applications. So you have inevitably a business user, let’s call him an analyst, involved in the process, you have a data engineer and you have a data scientist. The challenge has always been that these individuals work in environments of their choice. For example, a business user would work in Alteryx and use a low-code/no-code way of creating an analytical operation. A data scientist would probably be in a Jupyter Notebook or be writing a Python code. And the data engineer would be in dbt or writing SQL statements.
And because they all operate in their own environments, the mode of collaboration was asynchronous, it would be Slack or it would be e-mail or it would a variety of other collaboration sources, and that’s not really optimal. It’s time-consuming, it’s expensive and it’s error prone. We believe, with generative AI, we can enable unprecedented levels of collaboration between these people. We call that multi-modal.
And by helping our company and our customers embrace generative AI, we are allowing them to write instructions that’s in English to create analytical operations, that could then get handed over to a data scientist who would open that same instruction set in a Jupyter Notebook, for example, and start to work against that. And he or she could make modifications in Python, which would then kind of get picked up by the organizing workflow, and that could get handed off to a data engineer who would work in their tool of choice in SQL and make changes. So that back and forth between these personas, we think that unlocks a massive amount of innovation and really simplifies the lives of everybody involved. We are also allowing – helping our customers embrace generative AI more responsibly. So we’ve been training large language models on analytical operations. And our core strength is the millions of analytical operations that we’ve enabled over the course of the past couple of decades. So being able to train large language models on those analytical operations means we can allow every customer to customize generative AI for their analytical processes and their needs. We call that our AI Studio product. And both of these things that I was talking about, the ability to write instructions in English for workflows and the ability to customize large language models for analytical operations are going to be available for our customers later this year.
Mike Cikos
Awesome. It’s awesome. And I think another thing I’d like to highlight here, too, we’ve actually heard from a couple of companies earlier today. But just with respect to GenAI, there’s organizations that are putting in place, let’s say, I think the quote was a store chamber, right? If you think about like there’s a new management piece to the procurement process that’s evaluating this from a security standpoint. We’re making sure that everything is being properly implemented. We can explain why the model is giving the answer it’s giving. So in that capacity, is that something that you would agree with that you’re seeing? And then I guess the derivative is that in any capacity lengthening sales cycles as a result?
Suresh Vittal
Yes. So we’re certainly seeing customers being thoughtful and – but at the same time, progressive about how they embrace generative AI technologies. And customers differ. So it’s kind of hard to say one size fits all. But what we’ve seen from our customer base is a lot of interest in piloting some of these capabilities and really a lot of receptivity to talking to us about it because, guess what, generative AI systems and applications need well-prepared, well-formed data with millions, in some cases, billions of variables available to train, right? That means this whole journey of data preparation, blending, transformation, advanced analytics is something that they already trusted Alteryx to do. So they are receptive to having those conversations with us.
And then you also kind of hear customers kind of struggling between the variety of ecosystems that they’re dealing with. So would that slow things down? I certainly think it behooves every customer to be thoughtful about what technologies they embrace. What we do at Alteryx is going to give them that level of insurance that regardless of which foundational models or which hyperscaler they choose to work with, Alteryx is compatible with that.
Mike Cikos
Awesome. Okay. And I think you kind of touched on this a little bit in your response just now as well. But like, again, it seems like the market is still very much in an exploratory phase. You were talking about pilots, right, versus some of these moving into production. And so the question is like, what are you seeing as far as workloads moving into production on this front? Is it still very early? Or is there increased appetite or more workloads starting to move towards the finish line? I guess, how would you – if we were looking for a status update, how would you rank that?
Suresh Vittal
Yes, I think it’s – I think we’re still in the, I’d say, the first inning or you could even argue, if you’re using the baseball analogy, they’re still warming up, if you will, they’re throwing in the bullpen. I think that will continue for a little while longer because the ramifications are pretty enormous of getting it wrong, right? And so I think they want to – rightly, our customers want to do their diligence. They want to run pilots. They want to run hackathons. They start to kind of identify use cases where the application is relatively lower from a risk standpoint. And they start running these things.
So I expect it’s going to be that way for a big part of ‘24 as well. Before you start to see kind of mass adoption. So that’s kind of how I view this evolving. And not dissimilar, Mike, as you’ll recall from how cloud made its way into the enterprise or how mobile made its way into the enterprise or how social made its way into the enterprise. Rightly, CIOs, Chief Information Officers, are cautious because company data is their prized possession and they want to make sure they’re deploying these models the right way.
Mike Cikos
Got it. Got it. If I shift gears for a second over to ELAs, right? And so a lot has been made about this contract vehicle, what it permits as far as burst capacity, ability for customers to streamline adoption and really spread those analytics use cases throughout the organization, especially since it is a strategic initiative for a lot of organizations. The question I have with respect to ELAs is can you give us an idea how much of the customer base at this point has adopted, like where are we in that ELA penetration? And then secondly, I know that we’ve gotten data points in the past, but can you talk to how many of your customers who are on ELAs are actually they’re actively using that burst feature that you guys promote?
Kevin Rubin
Yes. Thanks, Mike. So maybe just backing up a little bit, for those that aren’t as familiar. So ELAs for us is a pricing and packaging vehicle. It allows customers in a very frictionless elegant way to purchase a large population of our technology, right? We have on-prem ELAs, and we have cloud-based ELAs. The context being we want to give our customers the opportunity to consume all of the technology that we make available, make it easy to purchase. They are bound by quantities of licenses, they’re bound by price and they’re bound by time.
So they’re not limited in use. They’re not forever and always, they’re just simply a packaging mechanism that we think is attractive to the customers. And certainly, it’s an easier method for us. I’d also like to mention they include customer success. So we ensure that we’re giving white glove treatment to our customers to help them in their analytic journeys and maximizing the use of Alteryx technology. The other feature that are included in ELAs that you’re alluding to is what we call burst licensing, which is in effect, we build these ELA bundles around numbers of designers. And so let’s say you’re buying a 300-seat ELA, right? That’s going to have some designer licenses, it’s going to have – it’s going to have 300 designer licenses, it’s going to have a concentration of servers. It’s going to have intelligence suite, customer success, as I mentioned, and really designed to optimize for 300 seats.
We will then also allow that customer to get another up to 150 designer licenses that is no cost to them that they can use for the very first year of their ELA. If it’s a 1-year ELA, then it just expires with the expiration of the ELA contract. If it’s a 3-year ELA, they get it for the first year. And it’s designed to allow them to really explore other areas within their organization where Alteryx can be impactful. It’s proof-of-concepts, it’s use case – proving out a use case, it’s going into other departments and identifying opportunities for customers to use it.
We rolled out this particular version of ELAs exactly 2 years ago. So Q3 of ‘21 was the initial rollout. We’ve seen continued momentum in customers electing ELAs. Not every customer is interested in a prepackaged set of technologies, some do like to very bespoke, select. This is how many designers I want, servers, etcetera. So it’s not the majority of the customer base. It still is a minority. But it is a much more frictionless experience for the customer. They don’t have to pick up a pick list and say these are the 27 different things I want. They have access to it all. They’re attractively priced. So when you think about those conversations around price per user, it tends to get quite attractive given everything that’s in the bundle.
And we’ve been really pleased with the performance. So Q3 was a significant quarter in terms of ELA as we more than doubled the number we sold on a year-over-year basis. Those may recall that Q4 of last year was the largest number of ELAs that we had sold in a given quarter. And so those are coming due and renewing in this Q4. So certainly, there’s some optimism and encouragement that we have this large population ELAs expiring and coming due. We do see customers using the burst, and that’s a great indication and future opportunity for expansion. It is a different environment today, and customers are cautious in terms of where and how they use software, and this is just not an easy macro. That being said, I think the ELAs have provided a nice vehicle for customers to consumer technology and do so in an affordable way.
Mike Cikos
Awesome. Awesome. And I did want to pause it like we – I’ll come back to that other question. Sorry, just put it in the bank. But first, it makes sense intuitive to me why someone would adopt an ELA, right? It does. So if I wanted to take the other side of the coin, and I know you were talking about customers might prefer like a more bespoke approach, but beyond that, like why would you not want to adopt an ELA? Can you help me think through that?
Kevin Rubin
Yes. So it could sit on either side of the pendulum, right? You could have a smaller customer with an estate of technology and the concentrations of the different products they have is what it is. And their willingness to take the ELA in the prepackage bundle, to them may be perceived as buying more than they need. And I’ll give you an example, right? They may have the large estate of designers and a small estate of servers, right? Or by the way, it could be the other way around, right? Depending on the use cases, we’ve seen environments where we’ve seen a heavier component of servers to designers depending on the use cases, and it may just not fit the ELA, right?
The whole purpose of the ELA is it’s prepackaged, and it’s a generalist package, if you will. It’s not intended to be customized. So you could be sitting there. You could also be on the other side of the fence where you have a massive estate of designer licenses and different concentrations of some of those other products. And there’s a sense that I’m either not buying enough and I’ve got to end up adding on any way or I’m getting something that is in a different configuration than I would have naturally selected myself, and they just opt to put that together. So there are – I mean there are natural reasons why customers are doing it. In this environment, I think there is sensitivity from CIOs and budget owners to make sure that what they’re buying fits their particular use and that they’re not being perceived as buying more or two of different concentrations of technology than what is needed for their business.
Mike Cikos
Got it. So I guess to build on that, I’d be interested to hear, it doesn’t sound like the ELA is going to be built for everyone, right? So over the long-term, and you can choose like, let’s say, it’s a 5-year time horizon as an example, or even longer, if you will, but like that ELA penetration of your customer base, is that expected to get to 50%? Is it closer to – like where is – have you guys gone through that process as far as thinking about what that ultimate penetration could be?
Kevin Rubin
Well, look, if Paula was joining us and you asked her, she would say to the customer to buy an ELA, right?
Mike Cikos
Yes.
Kevin Rubin
It’s the least friction path to getting customers to adopt and consume the largest quantity of our technology. So, it is not inconceivable to believe that it becomes more and more appealing and part of the motion, right. Part of it is also making sure that the sellers understand how to properly position the ELA for success as well. And we have seen continued success in that regard as well. So, I do think over time, customers that are going to purchase them, I think really appreciate the ease of purchasing and consuming. The sellers that are selling it certainly appreciate it. So, it is a heavy focus of us going forward. That being said, we are very customer-first and to the extent that a particular customer wants a bespoke set of technology, we are not going to, at the risk of customers’ decision-making, offer them something that’s inconsistent with what they are asking.
Mike Cikos
Got it. And this – this is the other question I was kind of holding in the wings, but it kind of builds on exactly what you were talking about. And so one of the criticisms that I have heard from people if they want to put together like maybe a bare thesis around the ELAs is like, hey, in year one, I am going to buy, let’s use the numbers that you had earlier, so it was 300 designer seats under the ELA, I will get 150 to burst. And if you find success with that ELA, the next year when you renew, you are not looking to go to 450, you are probably looking to go to, let’s say, 600. And so you go to 600 and you will get 50% burst and that will be 300 incremental. And now you are at 900 in aggregate. And if you play that out over the next year or something like that, you might find, hey, we were at 600 from the 450, we realized, maybe we only actually need more like 500 in total. And so there was – there is a concern that customers might have over-indexed or over adopted, which I know is not what your job is, your job is to make sure that customers are finding value, but I am just curious how you would refute that if someone who was posing it to you as far as concerns on over adoption from customers with respect to the ELA vehicle?
Kevin Rubin
Well, there is a few things I would point out. First of all, we generally don’t sell ELAs to new customers. So, they are generally positioned to an existing customer who is looking to accelerate their use of Alteryx. And I say that because it’s important. At the time we actually sell the ELA, we tend to see a nice increase in their state, right, from an ACV perspective. You just look at ARR pre-ELA, post-ELA. The ELA is a nice step-up in terms of the ARR of that particular customer. So, you get a benefit by moving them, generally speaking, to the ELA because they are buying more technology, they are consuming more. With the annual cycle of the burst, if we really want to play this out and a customer goes from 300 to 600, to 900, however that journey looks like, you are talking about two or three cycles before a customer may start, in this hypothetical example, optimizing use. So, we have taken a customer – let’s assume that these numbers are accurate, we probably took them for 20 – 25 to 50 licenses to hundreds of licenses to nearly 1,000 licenses over a 2-year or 3-year period to find out that maybe they need something less than 900 licenses. That’s still to me, feels like an incredible journey and an expansion opportunity with that particular customer. The other thing just to keep in mind, when we think about serviceable or addressable market in these larger organizations, we are talking about tens of thousands of employees in these large organizations that generally are data savvy and could be potential users. We are not talking about hundreds. And so I would like to believe that as you get into those conversations, there is so much opportunity and white space in these accounts. If a particular department determines, I only need a certain number of licenses, there are another dozen departments that have yet to consume and/or enjoy, and that’s what our job is, right. We have got customer success resources, we have got value engineering resources, we have got sales resources and their job, we have got SMEs. Their jobs are to go and make sure that those organizations and those departments we have yet to expand into are fully aware of what the offering is. So, while I can conceptually acknowledge that there may be a point at which a customer continues up this journey and decides they want to stop, if you look at where they have started, I mean that’s a tremendous opportunity path for us nonetheless. And then it’s our job to identify those new people and those new use cases. And Suresh’s job to put out amazing technology, as he has to attract those folks, and so I don’t know, it seems like it’s still a net positive journey, and we are still meaningfully growing that hypothetical customer account.
Mike Cikos
Agreed. And I also would argue that in this environment where everyone is more budget constrained, people are very much focused on what they are adopting in the first place, right? So, given the amount of scar tissue that’s been built up with these organizations and their management teams, the hope is that, we don’t get into that for quite some time, especially as it seems like, at the very least, it would be several burst renewal cycles away from where we are today.
Kevin Rubin
I would argue not to be the case.
Mike Cikos
Question for you just because I know the renewals base comes up frequently with investors as well. So, calendar ‘23, larger renewal base versus calendar ‘22, is there any way we should think about the calendar ‘24 renewal base? Again, it feels like with contract duration kind of stabilizing around 1.5 years that the size of the renewal base continues to grow in some capacity, but interested if you could shed any light on that.
Kevin Rubin
Yes. We obviously haven’t provided any 2024 guidance per se. I think directionally speaking, you are thinking about it correctly. We have had stable contract duration now going on almost 3 years. I would expect that from a – just from a general proportional basis that we are going to see similar dynamics as we go forward as well. But a lot of that will be dependent on Q4 as well. I mean Q4 is our seasonally strongest – largest population of renewals and strongest renewal quarter.
Mike Cikos
Got it. And probably time for maybe one or two more questions as we are bumping up against time here. But first, what’s interesting is I know that you guys have demonstrated this remarkably consistent level when thinking about the NRR for total customers north of $250,000 in ARR, right? And it’s – I would just love to get a sense for how it’s been so consistent. Is it a function of, hey, these organizations are that large and there is that much opportunity to expand in their state? Is it maybe that white glove service that we were talking about with the ELAs and probably focus from Alteryx on those large organizations? What are some of the puts and takes we should think about in the context of that very strong NRR at those levels?
Kevin Rubin
I think it’s a variety of things. So, first of all, in most cases, I mean I can’t think of a customer, and Suresh welcome your input here, either where we would say we are saturated, right. I mean we are very early in many of our largest customers from a penetration perspective. We have very large populations of users, but on a relative basis, these are very large organizations with large populations of potential users. So, your comment about the white space opportunity in these accounts, I think is important. We also continue to expand the suite of technology and the products that are available that we can sell into these organizations, which gives us a cross-sell/up-sell opportunity that candidly didn’t exist 2 years ago, right. It was really just we got to sell more seats and then we got a drop in a server. And today, we have an entire suite of the Alteryx Analytics Cloud platform that continues to grow on a week-by-week, month-by-month basis, providing more and more opportunities. And then lastly, I would call out customer success. I think in these largest accounts having dedicated white glove support to ensure that our customers understand everything that’s available to them that they have purchased and as well as to what hasn’t been purchased, and working with them closely in their analytics strategies, I think has proven to be a very successful formula.
Mike Cikos
Got it. And then maybe the last one, I know we are really up against it here, but would be on the go-to-market, right? I think management has done a good job on this most recent quarter talking about several initiatives that they use to increase visibility to the business model. And just wanted to understand, can you walk us through these developments? Where we are on driving these improvements on a more consistent basis or at least on a go-forward basis?
Kevin Rubin
Yes. It’s a lot to pack in here. But look, here is how I would, I guess maybe synthesize it. We knew coming out of Q2 that there were elements of the execution that we could do better, recognized that and took a number of steps in Q3 to address it. Things like enablement, coaching, linearity, making sure that the sales organization understands how to drive a value conversation, especially as budgets have continued to tighten, making sure that we are really aligned with our customers on their project needs and their project cadence, right. So, we are not getting ahead of ourselves. They are not getting ahead of us. Really understanding the quality of the pipeline, the deal progression, those conversations, We understand – many of the reps, I mean you can imagine, if you are a sales rep with 13 years experience, you have never worked in a challenging macro environment, right. Now, our more seasoned reps certainly have, but they haven’t done there so at Alteryx. And so there were some foundational execution issues that we have addressed. And then lastly, we have made two meaningful leadership changes that we talked about on the call. We changed out our Head of Americas. I am not representing that these two individuals impacted Q3 because they are recent adds. But nonetheless, as we think about going forward and continuing to refine on our execution, new Head of Americas, new Head of Partners and Channel. And I think that those two bring very unique stage-appropriate skill sets to the business that will help us iterate and evolve going forward as well.
Mike Cikos
Awesome. And I know we are at time. That was a great way to just slim it down. But Kevin and Suresh, and to the audience, thank you very much. I really do appreciate the time today.
Kevin Rubin
Our pleasure. Thanks Mike.
Suresh Vittal
Thank you.
For further details see:
Alteryx, Inc. (AYX) Needham Virtual Infrastructure, Data Analytics Software, & Cloud Communications Conference (Transcript)