Everybody, welcome to day one of the Citi Global TMT Conference. I'm Steve Enders, part of the software team here. With us for the first session, we have Expensify. Anu, I want to thank you so much for being here today.
Thank you for joining me. You guys started early.
Yeah, a little bit early here. Maybe just to get started, you know, for those who might be new to the story, can you just talk a little bit about Expensify, introduce the company, and tell us about your background and time there?
Absolutely. Expensify is an expense management platform. Let's restore.
Yeah.
Expensify is an expense management platform. We have corporate cards, travel, and of course the core product is expense. We primarily sell to businesses, but we have a range of customers, everyone from individuals looking to just manage their own finances to sole proprietors to small companies, all the way to really large Fortune 50 size companies. It was actually started on this, I want to say now may not sound as revolutionary, but at the time was that delivering a good product experience to employees is about as important, if not more, than delivering good product experience to the administrators of expense policies. No one really cared about the employee's experience, but we did.
We were the first to have a mobile app that had OCR capabilities that allowed employees on travel, on the go, to scan a receipt, never have to think about it again, and built in automation that factored in their policy settings and helped them do their expenses without spending their Friday evenings wrestling with paper receipts in Excel. Fast forward to the AI era, we're kind of doing the same thing where we built in an agentic layer on the app. We have brought chat into the app, which is front and center. The point of doing that is really to bring the conversations that happen around expenses to one place, contextualize it, and help admins mine it for intelligence so they can improve their policies and cut down on the time that it takes them to close their books.
That's a great intro for Expensify. Maybe just talk about, I think there's been a lot of changes with Expensify over the past few years, but yeah, maybe just kind of walk us through the journey of being a public company and where things are today and level set there a little bit.
Yeah, you asked about my background with the company as well, so maybe it's a good segue. I started at the company maybe to the day 10 years ago.
Wow.
I built our operations team from scratch. I built the card product. I've done a lot in terms of ops, product development. Now I'm primarily working on the sales and marketing side, leading those teams. When we went public, we were one of the few companies that only, we didn't raise much primary. We didn't go public because we needed capital per se. We wanted to create liquidity for our investors and our employees primarily. We were also one of the few companies that was cash positive, remains cash positive. The changes that being public brings about, generally speaking, is more compliance, more processes. We, as a company, always operated very process-driven. We are not really people-heavy. We are process-heavy, I'd like to say, in a good way. It didn't really change all that much in terms of how we operated.
Of course, there's a lot more in terms of storytelling and ups and downs with performance that you need to smooth out, for lack of a better word, for investors and the street. We're doing that, and we've gotten a lot of experience doing that because it's been four years since we went public. It's been a roller coaster, but only good times to come.
Sure. No, that's great too. That's good to hear. I want to touch on the sales and marketing a little bit, but I guess before we get into it, let's just talk about some of the newer kind of products that you've been introducing. I think over the past year, there's been, it seems like, a lot more, I guess, forward, you know, innovation.
For sure.
Can you just walk us through maybe what you've been releasing, what you've been working on, and how you're thinking about the opportunity for travel and the New Expensify platform and in cards and just what all that looks like?
Yeah, for sure. The reason we set out on building NewDot , which is basically our new and improved version of expense management, chat, and travel product, is because we kind of saw the need for a more mobile-focused, more real-time enabled, more conversational interface. That was something that we could do much better if we started over with a new product than if we tried to tape it on top of the existing product, which was really built for the last generation of internet SaaS applications, I'd say. What is the point of adopting NewDot versus staying on the legacy application? You have a very tightly integrated web and mobile product that performs much better no matter what your bandwidth, allows you to do things offline, and catches back up when it comes online. It's a much better app experience itself.
Then there's better functionality, which is a much more powerful search, a lot of admin features around managing data at scale that large companies and mid-market companies have long been asking for, integrated travel. There are upgrades, but also just the basic functions of the app itself are that much better. There are many reasons to adopt it. Of course, there's an agentic layer built on top of all of it, which, like I was talking about earlier, is constantly listening in a good way. Concierge, which is supercharged with agentic AI now, is also always listening to your conversation in order to intervene and help smooth the way where it makes sense. It's a much more real-time enabled application in every sense of the word. All of the unstructured data in the structured data's context, because unstructured data is the conversation, the structured data is the expense data.
The combination of both of it, using AI, leads to a lot of insights for companies that help them improve their policies, processes, identify who is most compliant, who is not compliant, and how to make their entire expense process smoother. That's really the vision for the product.
Okay. I guess when you roll out this new, the new chat forward solution, I understand that there's the agentic capabilities in there as well, which seems pretty interesting, pretty incredible to look at. When you think about the business or what that enables for customers, how do you kind of envision what that looks like or what's maybe been the feedback from customers so far?
Yeah, we're always, you know, this is something we've, our goal has always been to clear our customers' path to greatness, we like to say. It sounds a little bit big, but truly that's the goal. Nobody really wants to spend their time doing expenses. Everybody's really got other more exciting things to do, will have more exciting things to do if they have the time. That's really the goal of the product, to cut down on the amount of time an employee spends doing their expenses, booking travel, cut down on the amount of time accountants, controllers spend closing their books. We know accountants, finance teams don't really want to nag their employees to be compliant, to submit on time. Employees don't want to be nagged. We want to make a product that does not require these kinds of annoying behaviors out of any party in the company.
That continues to be the vision. There's just more and more tech these days that can help you, a SaaS product, accomplish that goal, and we're leaning more and more into that.
Okay, that's great to hear. Maybe shifting gears a little bit, I do want to touch on the go-to-market side of it, since we have you here.
Yeah.
I guess it seems like there's been maybe a bit of an evolution in terms of the go-to-market the past few years.
For sure.
Maybe you can just kind of talk about what you have done, what you've tried.
Yeah.
Talk about the F1 movie as well and put that in there. Maybe we can start there and we'll dig in.
You know, we have both a really good and a really challenging problem. The good part of it is we've invested a lot in brand marketing, and F1 is like the crowning glory of that. Why brand marketing? Because it is kind of a wide sweep, right? Like you communicate that you exist, your value proposition is, people get curious, they find, look for who you are. They may not be at the stage where they're making a buying decision today, but you keep on reiterating your presence and your leadership in the market, and when they are, they'll think of you. That strategy has generally worked. How do we know it's worked? We have probably the most robust inbound pipeline. We've invested in outbound before. I know we've talked about a period of time that we used SDRs to do a lot of outbound.
The ROI from any kind of outbound paid ads, none of it has really touched the volume of inbound that brand marketing has naturally pushed towards us. We know that brand marketing works, so we keep leaning back into it. The challenging part of the problem is, like I was saying in the very beginning, there's a diverse range of customers, and each customer, given their size and their particular problem, has a very different expectation of what the app should do for them. I think some studies said that you have like not even 10 seconds of a window when someone comes in and looks at your app and is trying to assess if it'll do what they need it to do for them.
Our challenging problem is really making the onboarding experience so fine-tuned that we can qualify the customer and then surface the exact onboarding content they need. Even if they don't finish onboarding, which almost nobody does in one session, we've convinced them that the product is for them, so they return. That's really been the focus of all of our go-to-market efforts. We've become much more, much more than before, data-driven, product testing-driven, and keep trying to fine-tune it. There's really no, when I started doing go-to-market as my area of focus, honestly, if you'd asked me, I would have rattled off a few things as silver bullets that I think such low-hanging fruit, like we're going to just do those things and it's going to dramatically shift. I've learned that it's really a slog.
You do a hundred things and you do the hundred things really well and it tips. That's our discipline now.
Okay. I guess maybe where are we in that fine-tuning? What have you found is working, and how do you think about what that means moving forward for some of the inbound?
Yeah.
Yeah, how do you drive that conversion?
Yeah, honestly, a lot of it is much, much, much better today versus even like three years ago, in the sense that we have a very strong degree of clarity on who the customer is and what problem they're trying to solve. We've become very good at surfacing them the right content. We are seeing a lot of that stickiness as a result. They keep coming back. Where the challenge really lies is a large amount of revenue is driven by a slightly larger customer, right? You have a huge number of small businesses, and those are really important because that's your word-of-mouth engine. That's where you get your revenue growth from. Then you have those small business, the upper end of the small business, middle market customer, where you get a, like once you onboard one of them, you immediately get a revenue bump.
They're stickier because they're not changing processes all the time. Now the market is kind of stiff in terms of competition in that segment. Everyone's competing on price. It's really a race to the bottom. Although that has changed a bit, most of our competitors used to be completely free three years ago, but they're not so much anymore. We hear rumblings about some of them being, in fact, almost expensive when they bundle travel and travel card and expense together, which is all great windows of opportunity for us. That's the challenge in that segment of the market. You need supremacy in both, really, to win. One is your word-of-mouth engine. It's a growth engine. The other is an immediate revenue boost. Long- term, maybe you can do it with just small businesses, but that's the short-term stickiness in revenue growth that I think the market's seeing.
Okay, and given those kind of dynamics and what you're seeing out there competitively, does that maybe change how you think about which areas to target? Does it change how you think about where you put the dollars and the strategy and put that to work to target that?
Now more than ever, I think it reinforces what has always been our strategy, which is focused on not just the mid-market and enterprise customers. Definitely deliver a product that mid-market and enterprise customers can use to their satisfaction. For a lot of our acquisition efforts, we're always focused sort of downstream, if you will.
Okay.
Now more than ever, I think that strategy is reinforced because here's a hugely untapped market. No one's really going after them, but it's difficult to acquire at scale. It's difficult to acquire using just a sales strategy with people because the unit economics are not there, and those customers don't want to be sold to like that. The trick is in figuring out and continuing to polish, like I said, the hundred things that are going to make that self-onboarding, support-based onboarding, lower ticket, lower touch onboarding more efficient. We stay focused on that. We've always really prioritized that, but now more than ever, I think the strategy is reinforced.
Okay. I guess with that kind of frame of view, I think we hear a lot from other companies utilizing AI and agentic capabilities to help with support, help smooth onboarding. Just maybe how are you leveraging agentic AI, you know, within the product and also internally as an organization?
Yeah, we start with go-to-market because we're just talking about it. We use it extensively. The way in which we use it is really exactly as I was previously describing. A customer comes in, they tell us a little bit about their size, their use case. Then they explore the product, which is really what they want to do. They don't want to talk to anybody, even if we're willing to spend that kind of resources in getting a lower ticket customer onboarded, which we have tried in the past. They don't really want to talk. They want to explore on their own. What we've done is send, we send every one of the signals that we get from a new lead exploring the product to the agentic layer. We have smart messaging capabilities built into it.
If you sort of looked at the QBO integration and then you went off to do something else, if you got an email or a chat about, okay, here is how you set that up. Here are the perks. It kind of brings you back to something that picked your interest. It's as simple as that. That's just one example. We also use AI extensively. Some of these leads are a little bit on the bigger end of the small business spectrum. They talk to a sales agent and it is important information that we could mine to know how to sell better to that segment. Maybe it's industry- specific, maybe it's size- specific.
Before, a lot of this was a sales manager to a salesperson, but nothing really existed to aggregate those kinds of intel to then make training across the Board better, to make sales scripts across the Board better. We're doing a lot of that as well. That's the go-to-market side. Again, on the product side, of course, we're doing a lot of testing, which is really how we're doing product marketing better. Across our operations teams, support teams, Concierge is now powered by much better agentic AI algorithms. We have an agentic AI learning or help documentation and will give you the perfect answer in less than five seconds. That's 24/7. That's huge in terms of support. A lot of the tasks that used to be somewhat onerous on the ops team have been more or less automated, except for the ones that require some sort of human judgment.
It just frees up people to do more and also gives us the ability to maintain our leanness in terms of human resources.
Sure. I do want to go into the product side a little bit more, but I guess what kind of efficiencies or ROI have you seen internally from doing that? To your point, I think Expensify has a pretty unique, maybe people model is the way to put it.
Yeah.
How do you kind of think about what that means for, you know, further headcount or how much more, you know, powerful and productive you can make the current employee base?
Yeah, you know, the reason we were able to always be so lean is because we always thought of all problems as how can we solve it with tech before we thought about how can we solve it with people. We always leaned more into what can we automate? We should automate that and not have people doing it. Now that question of what can we automate is just that much more powerful to answer yes to because there's just more resources given how good AI is and how much better it keeps getting. You've seen the cash flow forecasts that we have consistently revised up. We've been able to sort of maintain our, and you know, human costs at a SaaS company is the biggest expense. We've been able to use the fact that, you know, it's not just saving money, but it also is much more efficient.
We've been using that efficiency, which is really a double-edged sword, to also make our financials that much more performant.
Okay. That makes sense. Okay, let's talk a little bit about the product side.
Yeah.
With everything you've kind of come out with, New Expensify , I think looks really interesting. I think you touched on a little bit around how you're actually baking in agentic capabilities into that experience, but how do you kind of see that evolving further from here? What else can generative AI do for Expensify to help make the product that much more powerful?
We, I think from a product perspective, using AI, we've only just scratched the surface. Now a lot of users come in and they scan a receipt, and what we're seeing is there is already a lot of conversation now happening on new product within the context of an expense. Often what we see is somebody will ask, oh, I took this trip, can I expense business class? Or I don't have the receipt, is that okay? What we want to be able to do is teach the policy, and by that I mean the expense policy of a company, to the agentic AI layer and have Concierge answer them. It frees up admins from having to repeatedly answer the same questions over and over again because this agent works for them, basically.
It also surfaces for them aberrations in terms of expense behavior that it thinks they need to maybe take another look at, powering for a search into a natural language model. Right now, the way we all use powerful search capabilities is a bunch of different filters and tools. Some people will use it like that for a very long time, which is fine. We have a beautiful UX for them. We are going to make it also possible for you to ask it like you'd ask another human, and then it'll mine the data for you and give you insights. A lot of that sort of improvement is still in the pipeline, not yet live. We're excited for that.
Okay, that's interesting. I guess when you think about the go forward Expensify in the product side, just because what else could the platform look like? What new products could you potentially roll out? If you think about agentic in particular, are there other avenues of monetization that you could start to lean into moving forward?
For sure. The first one we did was cars, right? Like the first transactional revenue pipeline. Then we've now launched travel. They're both, I mean, the car has really propped up the financials in a very healthy way for a few quarters. It's already the proof is in the pudding. We have invoicing and bill pay, but we've got to think a little more in terms of is that initially the thought was that it's just going to be a word-of-mouth engine. Because when you invoice, when you pay bills, you're always introducing different companies to Expensify. That's in and of itself the way it pays. Maybe there is room there to look at transactional revenue opportunities as well. We've always talked about wanting to do payroll, which is probably not this year, but maybe late next year, thing. That brings with it its own revenue stream.
Ultimately, the idea is really to be like an Amazon Prime. You want to have one subscription. You don't want to think about 50 different things in your bill. Instead, you pay a price and that price is a rich value proposition to the company, but also a rich value proposition to you because you've pulled all of your back office financial operations into one product. Especially for the smaller end of the spectrum, ideally it's just chat-based. You have this agent in your pocket that can really help you do everything. You never need to interact with the UX if you don't want to. That's really the goal.
Okay. I do want to touch on travel. It seems like the initial release, the initial launch of that has been pretty encouraging and seeing pretty good demand there. Can you just talk about what you've seen so far from that product release?
Yeah, it's been, I mean, the year-over-year month-over-month growth is incredible. There's a huge waitlist.
Yeah.
We could go a lot faster, but typically with products like card and travel, you also have to be really careful about fraud. Doing things right is so much more important than just doing it really fast. We're just taking it slow in terms of releasing the dam, I guess. It's great that there is a huge pent-up demand, that there is a huge wait list, that we know how to onboard those customers onto travel better and better every day. We can do it economically. All very encouraging signs.
Okay. As you think about that future kind of rollout, what does the pace of that look like? You know, are there certain kind of checkpoints you need to see before you feel comfortable releasing it out to more of the wait list?
Yeah.
Yeah, how do you think about that?
We've basically created an entire team that is focused on cross-selling travel. We're certainly not, we were pacing it before, but it's all cylinders firing.
Yeah.
I forgot the expression. It's firing on all cylinders now. I think by the end of the year, we should be able to have soaked up all the demand. The question is really like, how do we sell it all together immediately? What happens generally is someone adopts core expense, they take six months, they may consider the card, then they take another. We want to be able to sell it as a package and communicate the value and power of the product immediately. That's maybe something we need to think more about.
Okay, I do want to touch on cross-sell.
Okay.
You know what that maybe looks like, because you do have a pretty big portfolio of solutions at this point. Is there, how do you think about building out a dedicated team for that? How do you think about how you can lean into that motion in particular to sell card and sell travel and get all that moving?
A lot of moving parts. I can't tell you exactly which way we'll end up going, but it's also kind of tactical. What we are doing right now is leaning hard into our account managers. They're the eyes and ears of the company in terms of existing customers. They know what their customers are doing. They know their pain points. They'd be the right person to pitch them these solutions that could improve their experience with Expensify. That's kind of where we're starting. A lot of them are, and we are running things like contests to get the creative juices flowing in terms of how we could be doing it better because those are learnings that we could incorporate. Down the line, if the demand keeps growing, which is such a great problem, we might use some of our sales people to do the actual selling.
The account managers might create the demand for it and then hand off to sales to sell. When it is sold, it'll go back to account managers to maintain. We might explore that, whichever works better for conversion. We don't have a perspective on one being better than the other per se.
Okay. I do want to touch on just maybe the macro and kind of what you're seeing out there in the last few minutes here.
Yeah.
What have you maybe seen from customers? How are they thinking about their own spend? How are you seeing the general demand and adoption of new software at this point?
Yeah, I think the biggest shift in our growth drivers, if you will, and we've talked about this a few times, has been the amount of expansion in usage among existing customers. If you take 2018, 2019, so pre-COVID, almost all of the growth really came from that expansion.
Okay.
New customers were always sort of compared to that, a drop in the bucket. Churn has been very steady for us, maybe spiked a little bit during COVID and then in 2021 when we had the price change. It's still sort of percolating, but it's smoothed out. By smoothed out, I mean it's sort of consistent for the business over the years. Now we're seeing a shift in terms of that expansion among existing customers is not as rich as it used to be. I think there's several reasons. Some of it we thought was macro, with a lot of companies letting a lot of their employees go. Some of it might be here to stay because with AI, more and more companies are becoming leaner. Maybe it's not going to be the biggest driver anymore, and that's a shift in macro technically.
New customer growth has been healthy, then was healthy briefly during COVID. It was a little stalled out, but it's healthy again. I think, I'm not sure what point of view to really have, and maybe jury's out on whether SaaS is dead, etc. We're not seeing that to be true based on just incoming leads, still very healthy. Ultimately, I think there's always two types of customers, right? The ones that want to use the product and the UX , and they'd like to do everything themselves. Even when you have tooltips and navigational help and videos, they just ignore all of it. I watch a lot of user sessions and just say, you're a new customer, maybe that video might, but they just want to explore on their own, do things on their own.
There are others that seek out support, seek out AI, seek out ways to do things without having to interact with the product. I think what we want to do is build a product that serves them both because we want them all.
Right.
That is really what has been the strategy and continues to be the strategy. The tactics change, but it's really still the same.
Okay, maybe in the last minute and a half here, or last couple of minutes here, you know, we can talk about just the longer term vision for Expensify. You know, let's say we're in five years from now, 2030 Citi Conference.
Yeah.
What would we be talking about then? What would be success for Expensify, and how do you kind of target the company strategically to get there?
For sure, we would have migrated and unified the products. We hope to have a large piece of the small business market, off of Excel, off of paper receipts onto our product, using a lot of the agentic capabilities to make their lives simpler because the easier use cases can do that the best. We hope to maintain, if not improve, our unit economics. Keep ourselves lean, really lean into intelligent automation that makes the company efficient but delivers the best user experience to our customers. Ultimately, we want to have a super app. We want our customers to use our product and no other because it solves all their needs as it relates to finance and back office capabilities. You know, then I guess we'll take over the world.
Yeah, just a little small step there to make that happen. Awesome. I think we're about at time. I want to thank you so much for joining us today and being the first session of the day here. Thank you so much.
Yeah, good luck.
All right, thank you. Thanks for everybody.