Well, welcome everybody to day two of the Citi TMT Conference. I'm Steve Enders, part of the software research team here. With us for the next session, we have the team from Asana, Dan Rogers and Aziz Megji. I want to thank you both for joining us today.
Thanks for having us.
Thanks for having us.
Yeah. Maybe just to start, Dan, I think it's been about a year since you took over as CEO. What have been maybe the biggest changes that you've made since you've come in? Aziz, maybe a similar question to you. Since you took over as CFO earlier this year, what have you evolved maybe in the strategy since then? Dan, start with you.
Yeah. Just winding back the clock, Asana was founded about 17 years ago. We were founded on the idea that as soon as you get more than two humans in a room, you have a collaboration challenge that you need to figure out who's doing what by when. Fast-forward to today, we became one of the most ubiquitous collaborative work management companies. 85% of the Fortune 500, about 300,000 companies around the world, are using Asana. But the situation I arrived in is that the modern team is no longer just humans that need to collaborate with each other, but humans and agents that need to collaborate with each other.
The first mission really was to reimagine Asana as the human agent operating system, to account for the fact that the collaboration that needs to happen now is going to be between agents and humans, and agents and agents. The agents in turn need that collaboration strata to figure out what's the agent doing next. In fact, does the agent understand the company's goals? Has the agent's work now handed over to the right human afterwards and to the next agent to pick up? This kind of task centricity that we found with human-human collaboration is also the same thing. Literally, it's the language of agents. They talk in the language of tasks. Anyone that's used ChatGPT, you'll see the kind of task hierarchy as it goes through how it's processing. We literally run on the task hierarchy.
The Work Graph that underpins Asana is really a relationship between tasks, projects, goals, and humans, and now agents. Reimagining Asana for the new era. That means 100% of our engineers are building against the human agent operating system. Instead of collaborative work management, now we call it Agentic Work Management, our core category, we're repioneering. Along with that, launching new products. Actually in the next couple of weeks is a big couple of weeks for us, because a lot of the things we've been innovating on come to market. You'll see not just one product, our core product, but five new products. We'll have new products for each of our buying centers, whether that's IT, whether that's software development teams, whether that's professional services and client delivery teams, whether that's process architects and process builders.
We really have a solution, and then marketing operations team, which is our home base. We have a solution for every team. That's on the product side. Then of course, go to market, we've been optimizing as well. We've brought in a new Chief Revenue Officer, and we've been very focused on sales plays, very focused on our pipeline discipline. You see it in our numbers. You see a nice uptick really every quarter since I've started, actually, both an uptick in our top-line growth and an uptick in our operating margin and an uptick in our NRR. It's been pretty consistent now on the, maybe it's three, four full quarters. Every single one has followed that same trend line. Obviously if we continue on that vein, it will be a great story.
Okay. Yeah. Let's keep the trend going. Aziz?
Yeah. When I started as CFO six months ago, it corresponded with the start of our fiscal year, and the world has dramatically changed since then. For me, it's really around how do we allocate our resources to the highest leverage areas to help drive this operating system that we're building and accelerate our growth. We've made a tremendous amount of investments that maybe we didn't anticipate in the planning cycle nine months ago around how do we invest in our R&D capabilities to build these new products? How do we bring on additional innovation? We acquired StackAI and absorbed the dilution, and that's been a really exciting journey. Early days, but we're seeing great proof points. How do we seed our customers with our AI capabilities? We've seen really strong momentum with our AI products.
Now we want to get them out to our full customer base. We're investing ahead of the financial benefits, so how do we absorb that? We've been able to have this rapid investment that will help drive future growth acceleration and also adhere to the financial commitments we made at the start of the year. That's been a real focus of mine, is really around resource allocation, ensuring we're placing our bets in the highest leverage areas, and adhering to our financial commitments while we do that.
Got it. That's great to hear. I do want to dig in a little bit into the guidance, and then we'll dive much deeper on the product side. Maybe we'll just talk through some of the moving parts here. I think pretty good beat for the quarter. Decided not to flow through the entire beat. Can you just maybe just walk through what went into the decision-making process for that and maybe think through some of the moving pieces on the outlook?
Yeah. As Dan kind of alluded, we had a strong quarter. We grew 10% year-over-year. It was our second straight quarter of growth acceleration. All the KPIs for the quarter were very strong, especially the NRR. Every reported cohort improved. As we thought about the beat, we beat our midpoint of our guidance by $2.4 million. We flowed through to the full year about $1.5 million. As we move to Agentic Work Management, which I'm sure you're going to unpack, there are some revenue recognition timing push outs that are associated with that. As part of our base becomes consumption, part of that revenue gets pushed out into future quarters. It's really only a Q3 phenomenon where we see that push out, because in subsequent quarters, you have the pull in from the previous quarter. That was about a $1.2 million impact.
If you think about that together, $2.4 million beat, $1.5 million flow through of the beat, a $1.2 million impact. In absence of that impact, we would've been rolling the full beat and then slightly raising from there. That's how we think of the puts and takes. As we transition to more of a consumption-based model in both the core package with AWM and then with teammates as companies expand with us, we're going to learn a lot about consumption. We're going to learn about how quickly customers get from adopting these seeded credits to paid consumption. We're going to learn about consumption patterns and ramp. Right now, with the data we have, we wanted to be a little bit more prudent in how we factor that into the guide, and so that was reflected in how we guided as well.
Okay. That makes sense. Last question from me on guide, and then we're going to get into AWM. Just on the product-led growth side of the business, that's been a, I think, recurring kind of headwind to it or to the overall Asana growth. How do you think through the headwind of that moving forward? What are the things that you feel like you can tweak and you can control, at least to a degree, to try to mitigate that headwind? Then, I guess lastly, just on the guidance side of it does seem like the assumptions on that downticked, but you again flowed through the guide, which probably means that something offset that. What was the positives that you kind of saw here, too?
Well, I'll maybe hit the business piece, and then you can hit some of the financial-
Yeah
Around it. From a business perspective, I'll say again, one of the beautiful things about Asana is a lot of our customers want to interact with us fully digitally or partially digitally. It is a real advantage to have a PLG motion. Many of our existing large enterprises, as an example, come in through the door of having a trial experience, getting started digitally. Whilst we have that advantage, it is true that the top of the funnel is changing constantly. Most recently, we've found that the top of the funnel has a lot of tire kickers who are not going to convert into paying customers, not going to enjoy the full richness of the product experience. The challenge really is how do we fill the funnel with more of our ICP?
Sometimes I think of it as some kind of entropy effect where all systems tend to complexity. It's probably the same with all funnels. All funnels tend to getting clogged up over time, and so you have to put a lot of concerted effort into getting back to the core verticals that we're focused on, the core personas that we're focused on. We made the pretty dramatic decision to change 100% of our marketing allocation to target verticals and target personas, instead of some of the generic programs that you might do on paid search, et cetera. That's a big change. The top of the funnel mix should change quite dramatically. When they arrive, we've also done a lot on the product experience.
Instead of arriving now to just collaborative work management, they'll arrive to Agentic Work Management, which includes teammates, but they'll also arrive to Asana Client Management, which is for those teams that are looking to do client delivery. That's about almost a third of our customers, actually.
They'll find that there's actually a dedicated product experience. The surface area of, I'd say, getting the right product to the right folks has just increased fairly significantly. We hope that that turn of the dial will change that profile quite significantly.
Got it.
Yeah, and then how we've factored in the guidance, if you think about the quarter, there's a couple things that are really working well. One is the strength in the enterprise. If you look at the metrics that kind of reflect our enterprise business, so our CRPO, which is correlated to the bookings that we see in enterprise. If you back out the large customer that we renewed about a year ago, that growth accelerated from 7% last quarter to 11%. If you look at our 100,000+ customers on a customer count and logo basis, those accelerated to 16% growth year-over-year from 12% last quarter. We now see 25% of our 100,000 customers adopting AI, which is our AI products, which has led to strong improvement in NRR in the 100,000+ cohort. That cohort improved their NRR from 96% - 98%. So enterprise, really strong.
We're seeing strong momentum in our AI products. 25% of our net new ARR was with our AI products this quarter. That is what's driving kind of the flow-through of the beat, and the revenue growth equation. Now, on the flip side to your PLG question, we are seeing the impact that we called out at the start of the year, that we were going to see two points of ARR headwind from the PLG business. That has increased in Q2. Now it's around 2.5% impact. And the flow-through of that AR impact, we're seeing more pronounced on a revenue basis in Q3 and Q4. So the impact to our revenue growth is about 1% in Q3 and about 1.5% in Q4.
If you take those three things together, strong enterprise, strong AI product adoption, which is improving NRR in our larger cohorts, and then this continued headwind on self-serve or PLG, which you're seeing actually in the NRR of our overall PLG base, which is muting a bit the improvements in the NRR in our larger customers. That's kind of the puts and takes and, as Dan walked through the initiatives, it's a big focus of ours to get that business re-accelerated, because as we do, it's an accelerant to both our revenue growth and our NRR.
Okay. That is very clear there. All right, shifting gears now to AWM. I think it really seems pretty interesting what you are doing here. Maybe we can dig a little bit into what is generally new in the product with AWM. How does it maybe evolve what you were doing before with CWM, instead of it just being a repackaging or a renaming of the prior solution set?
Yeah, great. Maybe I will talk about what is happening in the real world. In the real world, there is this AI productivity gap. AI productivity gap is people are experiencing a lot of personal productivity increase with working with the chat agents, with the LLMs. But companies in general have not really translated that into their workflows, into their core productivity, and so company footprint of AI tends to be fairly low or single-player mode of multiple people now producing their documents better or doing their coding better.
That maketh not a business process. We said, how do we more deeply ingrain agents into what we call the agentic enterprise? The answer is through Agentic Work Management. So what is Agentic Work Management? Well, the first idea is why do not we prepackage up some agents that can work alongside humans, but unlike these single-player agents, let us make them multiplayer from the get-go, and let us make them, instead of you having to buy these things or find these things, let us make them raise their hands as you are doing your work. So 30 pre-built teammates.
So teammates across, of course, we have had 17 years of history of what people are trying to do at work, so 30 teammates that actually cover the large part of what people are trying to do at work, and we have pre-skilled them, pre-trained them, and pre-contexted them on your Work Graph data. So what does that look like? Let us say in marketing, where we have five of these pre-built agents, we do have a campaign planner in marketing that is pre-skilled and pre-trained on what it takes to build a marketing campaign. It literally knows that every marketing campaign has these 16 steps, and you are probably going to need to have brand approvals, and you are probably going to need to link into your core brand assets, and you are probably going to.
So what it will do is if you are in Asana and you wake up in just a couple of weeks' time, you will open up Asana, you might be working on a campaign, you might be writing a task. This thing is going to volunteer itself and say, "Hey, I am your campaign manager agent. I got you." What the heck is this thing? Let me click on this thing.
Yeah, I know so much about what you're trying to do already because I've scanned the Work Graph and figured out who you work with, how you do this. By the way, I've actually preordained from the people that have trained me what maketh a great campaign. Here's what I think you're going to need to do next. Are you game? Sure. You'll start working with the campaign planner, and it will then in turn say, "I notice that these are the 16 people that you work with on campaigns. Do you want me to kind of tag them in too?" Okay, great. The next thing you know, we're all working with that agent. This is a completely different paradigm shift.
We have millions of users, and in a couple of weeks' time, millions of users are going to each have two, three, four, five, six agents that they're working alongside that have volunteered themselves to help. So literally, we are going to go across the chasm in a massive leap to seed agents in the modern enterprise. Prepackaged, pre-skilled, pre-trained, operating just like humans, and because they're in the same context as the rest of the humans, you can fully govern them, you can fully audit them, and you can fully scope them to work just as the rest of the human teammates. So that's thing one.
Again, to your point, this is not, "Oh, we've got some new AI feature. We've got some new virtual chat assistant." This is actually germane to how you work. It's deeply embedded in your workflow, so it is a category-changing shift.
The second thing is we have developed a chief of staff that we call Dash, and Dash will also be rather pleasing to any of our users because Dash will, again, be fully contexted in the Work Graph, be fully contexted in the teammates that can help, and just be able to pull not just on everything that's in Asana, but also everything that's in your email and everything that's in your chat with Slack, and everything that's in your calendar, and then actually infer, oh, and everything that's in your meetings from maybe your meeting transcripts, and infer the tasks that need to happen next and what you're due to do. Actually, any of the tasks that you currently are working on, what new information that's in that peripheral system you need to be aware of that can be enriched in that task.
Again, it's going to be a huge leap forward in what one can imagine an AI chief of staff should do. Again, really pointing to Aziz's point about we've been busy.
From an R&D perspective, I really can't describe how the velocity of our engineering teams is probably 2.5x in the last 12 months.
Wow.
We are shipping stuff that's game-changing. I think the big reveal is just a couple of weeks away, and I think people are going to be very surprised about the new Asana and what we've been up to. Then workflows in general, both through our acquisition of Stack and then also through this thing called AI Studio, the ability to visualize processes and actually have those processes connect not just to agents, but also to third-party systems. If this, then this. For example, Citi's built their own virtual assistant called Sky. Looks amazing. Got a preview of it last night.
But in reality, when a customer has a discussion with Sky, it's probably going to end up forking into one of five paths that's going to need some other workflow that falls out of it or some other team that needs to get involved, and that's really where Stack comes in. Visualizing workflow, connecting to other systems. It'll end up being the tapestry for the new enterprise is the idea.
Okay. It is really interesting, I think, how it is evolving and how you are thinking about what the future of Asana looks like. I guess I maybe want to ask about the diffusion of AI into the enterprise and how you see this spreading. How much are you thinking about this being something that just gets surfaced naturally within the product set, and that is what is going to drive the end user to start using this, versus this being more of a sales-driven push and really trying to just get it in front of all the customers and needing to train the go-to-market team on that?
Yeah. I spend a lot of time with CIOs, and I talk to CIOs about this great AI productivity gap and what is the reason behind it. Yes, there are security concerns and governance concerns, but actually, the biggest one is discoverability, is that teams just don't know how to use various agents and which agents they should be using. There is a huge initial productivity or cold start issue, which is don't know how to get started with said agents, don't know how to find them, don't know how to get started with them. That is the challenge we wanted to solve, and very deliberately. We have had six months where we had these things called Asana AI Teammates that we were selling as an add-on.
Not discoverable and not germane to the core experience. That is fine. We got, as Aziz was saying, a 25% attach rate of our AI products. But what would a 100% attach rate look like? That is really where in-product notification, in-product nudges, really being very contexted as someone is trying to work and something that will literally help them. Trying to do that in an unobtrusive way, I think we may have nailed it. I suspect that most people will want to try these things. We have also been very thoughtful about making sure that there is not a billing event when they try those things. Right? So how much do we give to get them to a point where they are delighted has been almost like a scientific endeavor, and I think we have found the right amount there.
The unit we are using is requests, so we are going to seed requests for every user, so that really you should feel like, why not? Why wouldn't I give this agent a go? It is going to pass my company's security requirements. It is going to pass my governance requirement. It is there waiting to do work at my bidding, and it seems to know a heck of a lot about what I am trying to do. Let's give it a whirl. Because we have seeded five requests per user per month, we think that is enough for them to get to a point of joy and saying, "Okay, we have got to get more of these teammates. I have got to use this more. It is germane to my work."
Yep. Okay. When you think about the monetization potential of this down the line, understanding you are trying to drive usage and seeding it out to customers, I guess, what is the next step to when we start to see this hit revenue in the monetization path for you all?
Yeah. What we have seen with the adoption of Studio and AI teammates has given us a lot of confidence that getting this to a fuller base, as Dan said, reducing the friction and the barriers to adoption and engagement will help drive that consumption path a lot quicker. We have not defined, it is not in our guidance. We have been conservative and prudent how we have factored in. As we think about our model, we have grown with customers primarily through headcount changes. As they grow headcount, we grow. Now we have multiple ways, in addition to headcount changes, to grow with customers. We can grow with outcomes, and the outcomes we drive greater consumption and utilization of our products. We can grow by the work being done and the output we are creating.
I am sure you will touch upon it with our new agentic apps and expanding buying centers, we can now move to different departments with outcomes as well. It just enables these additional growth vectors that we have never had. We are seeing that with expansion. Our NRR improvement, a lot of that is being driven by expansion, driven by our AI products, and we are seeing that with renewal activity. We had a large customer that we called out in the quarter where they have less headcount than they did three years ago when they renewed with us. We piloted AI Studio and AI teammates with them several months ago. They saw value in one production workflow in a marketing application. They saved about 30% of time on creating content for marketing campaigns.
That ultimately led to a renewal where we were able to offset a downgrade of seats with Studio and teammates and create an expansion outcome. This is a multimillion-dollar TCV type deal where our AI products are half of that. Being able to expand with consumption and outcome-driven products, a year ago, that would have been a downgrade for us.
Now it is an expansion with an opportunity to drive a future expansion as they continue to leverage these products and drive more work through them. These are exciting new levers to our story that just add additional growth vectors in the long term.
Okay.
I will just say one other comment is, obviously we are pioneering from a product innovation perspective, but we are pioneering from a business model perspective as well. This notion of literally across millions of users, multiple agents volunteering themselves, that does not exist. How should one forecast that? Then we have got the conviction confidence to do this based on our, I would say our run up of the last four to five months, but we are taking a huge leap, and so it is important to be prudent and conservative. But yeah, I think we are also being very pioneering.
Okay. No, that is interesting. Maybe this is a good point to ask about net retention. You saw a two point jump this past quarter on the higher end of it. I guess, what do you need to see happen moving forward to get that number to 100% and potentially above that from here?
Yeah. If you think about our net retention as a rolling four-quarter metric, each of the quarters in that rolling four quarter are improving. We have had five straight quarters of improvement NRR. Even if we just kept our in-quarter NRR where it is today, you will see sequential improvement. To drive it to 100% or greater, which is the goal, it is really three things. We have two things that we are doing really well that continue to drive better net retention. One, we are seeing the gross retention part of that equation improve quarter over quarter. Our utilization of our underlying seats is improving, which that is highly correlated to net retention. That is number one on the gross retention. Number two, we are seeing the customers as they engage deeply with our AI products and drive better outcomes.
We're seeing better seat dynamics, both in retention and expansion. Our AI customers are actually expanding much faster than our overall base. Expansion is an important lever. A year ago, it was about pricing and seats as levers to expansion. Now we have five products we can expand with, both seats and consumption and outcomes. So that gives us just a lot more in our arsenal to expand with customers across additional buying centers. So that's the second. The third piece is back to the self-serve. What's driving down our net retention? If you actually back into what the in-quarter would've been for the 100,000 and 5,000, they look better than 98%. They're actually approaching 100%, and one of them is already above 100%. It's really that sub 5,000 cohort that's bringing that down to 97% on an aggregate.
All the things that Dan outlined are all aimed at improving the lifetime value of those customers, having them land at higher ACVs, expand at greater rates, driving more value for them through having more personalized teammates and products, which will improve the retention. So it's continuing gross retention, much more to sell, which will help accelerate expansion. As we get that PLG and self-service piece, that NRR improvement will be a big tailwind to the NRR.
Okay. That's good to hear. I think we only have a few minutes left here, but I want to make sure that we touch on some of the other new products you have, like Client Management, Service Management, and Command. I guess maybe first of all, what went into the decision-making to create a specific solution like those, and how do you think about what that means in terms of augmenting the business or expanding the TAM that Asana can go after?
Yeah. First, maybe you have to understand a bit about my own history to understand that question. I guess my formative years were at ServiceNow and then at Rubrik, and I'd say maybe get a page in the book or a chapter in the book at both of those companies on how you go from single product to multi-product.
I was running a lot of the strategy teams at ServiceNow when I was CMO there as well. But really, I guess my main task was how do you go from being an IT service management company to being a multi-workflow platform? Similar to Rubrik, how do you go from being a data protection vendor to a multi-platform, multi-product data security platform? I'd say part of it is this is what I do, so that's maybe a bit of the history.
But specifically to Asana, it was obvious to me that we were already serving all of those buying centers. We were just doing so with a horizontal approach. So very nice correlation under the hood of, like, oh, wow, x percent of our teams that we're serving are IT teams, y percent are R&D teams, z percent are marketing teams, fu percent is operations team. It was a beautiful picture.
Then when you saw what they were trying to do in the platform, where they were trying to push the platform, what the edges were of where they were trying to go, there were specific problems they wanted us to solve. At a certain point, it's hard to do that in a horizontal canvas. We said, well, why don't we keep the horizontal canvas but be a lot more declarative from some of those vertical use cases? So that's what we've done. We've set about creating these vertical products, but because they still sit on the same operating system for human agent teams, the better together story is actually amazing. It is, yes, you can have a teammate that still works across all of these other departments.
Our angle in is not, oh, let's take out the existing IT products, let's take out the existing Asana Client Management products, let's take out the existing R&D type products like Atlassian, et cetera. Those might be byproducts, but the reality is we're going to better serve everyone this horizontal across the company by having better solutions for each. The solutions will be amazing because they will be designed for the here and now. We were very lucky because we began R&D on all of these efforts, let's say post the Codex Cursor revolution.
We got to build and imagine for the agentic future. Our approach to IT service management is as modern as it gets. It literally was imagined nine months ago and shipped today. With all of the, I'd say, underpinnings of what that could look like, what real-time knowledge bases could look like, what self-learning agents could look like. We believe no one has this kind of solution because they're all encumbered by the past way of solving it. We've literally solved it for the next generation. Every one of those products, I think will be quite mind-blowing when people see them.
Okay. Maybe this is a good, we're running up on time here, maybe a good time to plug the Work Innovation Summit coming up in a few weeks here. I guess, what should we expect coming out of that?
Aziz, do you want to take this? I feel like I've had a lot of airtime back.
It's more your domain, but-
Okay, I'll do it. So, yeah. Again, we have two Work Innovation Summits in London and New York, and they basically, for us, they become development milestones as well of things that we're going to release for each. And you could think of it as kind of summer release and winter release actually for us. And increasingly, we're going to start describing that to the outside world. So really, it's kind of our winter release or winter release preview. But all of these products will be in market by then, and so you'll have a lot of customer stories about real customers that are doing it, that we've been very coy on, and we've had to be very coy on because they've been in this kind of design partner process.
But you'll see a lot of big brand names that have been using these products and getting to these agentic outcomes, delivering real productivity baked into their workflows. And so I think it will be a rather grand unveiling of what it looks like to run a company that is an agentic enterprise.
Yep. Then we will have an investor and analyst fireside with Dan, myself, and our Chief Product Officer, Arnab Bose, to kind of elaborate on the innovations and the customer stories and this business model and product transformation that we're undergoing. So we're looking forward to that, and it should be some exciting announcements.
Perfect. Well, definitely looking forward to that. I think we're at time here, so we'll leave it there. But Dan, Aziz, I want to thank you so much for being here today.
Thank you for the great questions.
Thanks, Steve. Appreciate it.