Upwork Inc. (UPWK) Earnings Call Transcript & Summary

Good afternoon. I'm Josh Chen, business services analyst here at UBS. Very pleased today to have Upwork join us. They are an online marketplace connecting freelancers with SMB and enterprise clients globally. With us today from the company is Erica Gessert, CFO. And we'll have a fireside chat here. Feel free to raise your hand. You can send some questions in this iPad, and I'll pick them up as well. So -- but with that, Erica, great to have you at the conference again.

Thanks for having us -- thanks for having me, Josh.

Yes. Thanks for being here. So maybe to level set the audience about Upwork, could you start by giving a brief intro about the company and for anybody that may be less familiar and then we can dive into different topics from here.

Sure. Sounds good. Yes, Upwork is the world's human and AI work marketplace. We are a global marketplace that connects talent from all over the world with businesses of all sizes. We serve globally. We have over 130 job types on our platform and thousands of skills of all size companies.

Great. Great. So maybe to start off with the recent return to GSV growth. So could you frame for us the GSV trajectory over the past several years? And then why it was briefly negative, but then more importantly, how you've been able to return to positive GSV growth recently?

Yes, sure. Yes. Look, the past few years, particularly, I think, as companies exited the pandemic, there were many pandemic beneficiaries. We were included in that. And as companies came out of the -- as companies came out of the pandemic, I think the broader human capital industry really kind of suffered in that environment. And most of our kind of staffing industry competitors were down double digits from the volume point of view over the past few years. Upwork was relatively flat. And we kind of really worked over the past few years to invest in 3 key growth catalysts for our business, which we're now seeing an inflection point across all 3 of these areas. So there are really 3 pockets of very large TAM that we have access to that are really showing very strong growth characteristics right now. The first, of course, is the AI category of work itself. And kind of the AI opportunity for our business is kind of threefold. The first is growth of the AI category, which is accelerating on our platform. It's at a run rate of about $300 million of volume on our platform right now, growing at over 50% a year and accelerating. That's largely organic growth. And so we're making some investments now to even further catalyze the growth of that category. And so it's super exciting time there on the marketplace. The other part of AI growth is actually the front-end customer experience and the AI enablement of the hiring experience itself. So if you think about the hiring experience, it's really traditionally very friction filled, right? We're a marketplace. Our clients would come on to the marketplace, have to write up a job post, post the job, talent writes a proposal for the job and the client has to read through those proposals and then choose talent. Now all of that is done by AI. AI writes the job post, AI writes the job proposals, AI helps with the match all across. And now even in more recent quarters, and really in Q3, we've launched the ability for AI to do recruitment, to do interviewing and other things. So just in 2025 alone, we saw about $100 million GSV lift from these kind of experiences to reduce friction. And then the third opportunity from an AI point of view is actually a little bit longer horizon, but introduction of actual agentic work on the platform, which we've just introduced kind of benchmark work where developers can come and test their AI agents doing work on the platform as well. So there's a huge growth catalyst there. And then there's a second growth catalyst on the marketplace, which is the growth of the SMB category itself, which we can get into a little bit more.

Sure. On the AI topic, I guess, there's been discussions about positive negative effects. Clearly, you highlighted the GSV growth from AI recently. How do you think about any headwinds? And what's the aggregate impact of AI in your business, you think?

Yes. It's a really good important question because I think that we've -- businesses like ours that are in the human capital contingent labor industry have had an AI overhang over our stock and valuation, certainly since during my tenure in this business. And I think it's been really a misunderstanding of kind of the external markets in terms of how and where disruption happens in the labor market. And we have actually one of the best data sets out there on this topic because we have hundreds of job types on our platform and billions of dollars of volume every year, right? So we can see very clearly where there is substitution and where there's actually augmentation from AI and then even acceleration with these kind of net new job types. The places where we did see substitution, and we've talked about this a lot, is the biggest categories that saw substitution were in writing and translation. A few years ago, writing was a relatively big category on our platform. Now it's quite small. But even more so, where we saw substitution was we have to sort of a proxy in our business that we talk about when it comes to substitution. And it's really jobs that are $300 and less on the platform. So if you think about where AI can substitute human work, it's really in the smallest, most transactional types of work. Now our -- so a couple of years ago, the $300 and lower job types were about 5% of our volume. It's now about 3.5%. Our platform overall actually specializes in longer, more complex projects. And so our average GSV per client is actually about $5,000 and growing. It's growing sequentially every single quarter. And that is actually partly because of the growth of AI category itself, which is 3x the GSV per client that kind of our platform averages. And then also actually the kind of our further penetration into larger SMB customers who, by nature, do much more complex and longer-term projects.

Any questions from the audience so far?

I [indiscernible] either.

Yes. So -- and then maybe transitioning to the macro backdrop, which employment markets have been tough like you mentioned. So what's your view about where we are in the macro now and then kind of as we're heading right into 2026?

I mean, look, I would say the macro is not doing anybody in any favors right now. I think that inflationary characteristics and sort of persistent high interest rates are not great for particularly SMB spend. But that said, I think that we have really put in place the kind of growth catalysts that I described, AI, the growth of SMB, which we have really built out the products on our platform. We have a strategy where we're kind of tiering kind of our servicing on the platform and features on the platform in order to have greater service for these larger SMB customers. So if you actually think about the kind of history of the Upwork platform, it's really been kind of one size fits all and optimized largely for -- if you think about SMB customers, these are companies that are ranging from anywhere from like 1 to 9 employees all the way up to 100 to 200 employees. And the needs of those customer types are very, very different, right? So if you think about a 10-person employee company, they can go around on the platform, they need a software developer for a couple of months. They can just hire it's super easy, right? If you're in a 100-person company and you need a group of developers to build a long-term complex project, well, you probably need a real invoicing product like you need net 30 billing. You need the ability to build teams in an easy way where they can collaborate on platform. You need the ability to have different people interview these people, all these kinds of things, we didn't offer any of those types of products that are really custom-built for these larger businesses. We just launched all of those in the past year and really some of these features in the past couple of quarters. And this is now a premium take rate product that is also aimed at these much larger businesses and longer contract types, and we're starting to see some real acceleration here. In Q3, the GSV in this business plus category grew 36% year-over-year. That's also accelerating. And it's still very low penetration on our platform. Less than 5% of our GSV is in this kind of new tiering product. And so this is a super exciting catalyst for our business as well.

And what I'm hearing you say is that regardless of what the macro is, you feel like these catalysts can kind of continue to help you grow?

Yes. I think for us, we've really spent the last couple of years rebuilding the platform in these new profound ways that are truly different from anything we've done before. The third area actually is in the enterprise side of our business, which is separate from the marketplace and really focused on the very large Fortune 200 businesses in the world. Now for those who don't know the contingent labor market that well, these businesses -- I actually came from large enterprise. These businesses spend hundreds of millions of dollars in contingent labor a year. Most of the large enterprises, their kind of employment portfolio is, call it, 70-30 FTE to contingent labor or sometimes even more. And so there's hundreds of millions of dollars of opportunity in all these businesses. Previously, Upwork was really only offering one contract type to enterprises, which was an independent labor contract type. That meant that we only accessed about 10% of that enterprise wallet. And in Q3, we announced 2 acquisitions to kind of change that. We acquired a business called Ascen, which offers all the different contract types that we were offering before. And then another business called Bubty, which is actually a workforce management platform that can plug in API-ready into all these enterprise systems. So that's going to be a huge game changer for us. It opens up the other 90% of a $650 billion TAM enterprise market that just was not accessible to our business before. Now we're in the middle of integration. We're not quite ready to kind of launch all these new contracts, but that's coming in 2026. And that's another reason that we feel very strongly that we can kind of be macro agnostic as we penetrate these very large TAM opportunities.

Okay. That makes a lot of sense. Recently, you had an Analyst Day and you outlined some near- and medium-term targets. So maybe just like stepping back, what made you decide to put together an Analyst Day now? It really does seem like you feel like there's a little bit of an inflection?

We definitely do feel like there's an inflection point in this business. And I think we planned the Analyst Day very carefully for that reason. We've been hard at work for 2 years on all of these strategies. We actually -- if you go back and look, we -- at the end of 2024, we said that we were going to -- 2025 was going to be kind of a negative GSV growth year for us as we rebuilt a bunch of these strategies and these businesses. And we even -- we announced at the end of '24 that we're going to stop selling our legacy enterprise plans while we worked on this new enterprise strategy of acquiring these companies that could offer the full suite of products that these businesses needed. And so we really did take a step back, rebuilt and we are now seeing the benefits of that work across all 3 of the fronts that we have worked on, AI, SMB with Business+ and now the enterprise business. So it's a super, super exciting time for us. And if you look at our guidance, it actually implies kind of real acceleration through the next few years. And we feel really confident in that trajectory because we are seeing all -- the nice thing is all of the strategies that we've described are things that we are already seeing traction in our business today.

Sure. Yes. On guidance, you did guide to 2026 GSV growth of 4% to 6%, which is an acceleration from the....

Yes. Yes, that's right. So we just reported in Q3, 2% GSV growth for our business, and that's coming off of about -- actually 5 quarters of negative GSV growth. And so there really is an inflection here. And we've guided to 4% to 6% GSV growth for 2026 and then 7% to 9% for the next 3 years. And so we feel very confident in that trajectory. And the reason we see that acceleration is because it's both on the marketplace side because of the ongoing acceleration and ramp of AI and also the AI implementation on the platform, it takes time to implement these experiences from a customer experience point of view, inject AI into the platform from a recruiting point of view. We're just about to launch AI project management. These things take time to kind of build into the experience and then ramp. But every single -- one of the kind of AI experiences that we've launched so far, where we've reduced friction from the kind of historical experience, we've seen an inflection and increase in fill rate and other things on the platform that enable our growth. So we feel really confident of that ramp over time kind of in all 3 of these areas.

Okay. You talked about AI categories. So how do you define what's an AI GSV or not? And then from -- in terms of the guidance, what does it embed in terms of AI growth versus non-AI trajectory, I guess?

Yes. I mean we talk about a lot about AI on the platform because it is such an exciting time and such -- seeing so much acceleration right now. Now we are a very diverse platform, though, I think to your point. And we see growth across other categories, too. It's a very interesting time in our business because, like I said, we have such a great data set. But categories where I think people thought there was going to be AI disruption like design and creative, logo design, these are things that are accelerating. And actually, we see -- within our data, we can see categories that are actually getting augmented by AI tools. And so as like tools get better, those categories are accelerating. Video and animation is actually one of those that we -- some of the AI queries on our platform, AI video is kind of one of the highest that we see. And so we're seeing real acceleration in some of these kind of subcategories as well. So getting to your question on how do we define AI category. It is difficult. Our taxonomy in some ways can't keep up with it because of the use of AI tools. But when we talk about the AI category, the vast majority of that is today, as we characterize it, is what you would think of as like typical AI work. It's either AI infrastructure implementations, it's prompt engineering, data labeling, things like that, that are kind of underlying that growth. But there are also other -- we do try to identify other kind of AI-enabled job types that are somewhat in that category, although it's difficult for us to always identify them all in real time.

Sure. Do you need the non-AI categories to stabilize to hit your guidance for '26?

I mean many of the non-AI categories are growing nicely, like design, creative, legal, accounting are all actually very nice growers on our platform, sales and marketing. And so no, I mean, look, we don't need any change in our trajectory to hit our guidance for next year is how I'd characterize it. And I think some of the work that we're doing right now is on top like in terms of further catalyzing the AI categories, assuming some of that work really works, then that would be incremental.

Sure. Any questions from the audience? Okay. So -- and then maybe moving on to take rate. So implicit in your revenue guidance being higher than your GSV guidance is a slight improvement in take rate next year. So what's driving that?

So take rate, we have a tremendous amount of opportunity with take rate. The first most important kind of contextual piece here is that our take rate is actually quite -- still quite low compared to the rest of the industry. So our take rate all in is about just -- it's about 19.6%, just under 20%. A lot of industry comps are well into the kind of 30% take rate. So we do have room to grow here. The really nice thing about our take rate strategies that we're implementing right now is they're all very much pricing to value on the platform. So one of the biggest catalysts for us is Business+. It's this SMB-focused product that I described. This is a premium take rate product. So on the client side, our clients pay 5% take rate on our platform. Business+ is 10% on the client side. And it's also aimed at these much larger contract types, as I described. And so as Business+ grows, it will be a catalyst both to volume and to take rate at the same time, which is a very nice place to be. The other area that's really interesting for us is that we just started to experiment this year on a variable freelancer fee within the platform. And this is really going subcategory by subcategory and looking at the supply and demand dynamics of the platform. And in some places where we have a surplus of supply, we will increase take rate. And in other places, actually, we've experimented with reducing take rate in kind of lower supply categories in order to catalyze GSV. And so that's also going to be -- we're very early days. We've only experimented with a couple of categories on the platform. And so this is something that will be a really nice tailwind into 2026.

Okay. Okay. And then kind of going beyond 2026, your 3-year GSV target is 7% to 9%. So that implies a further kind of acceleration from here. So can you talk about what's driving that?

Yes. Look, I mean, first and foremost, the enterprise strategy that I've outlined with our enterprise business is called Lifted. And that strategy -- the game changer there is our legacy enterprise business had a very wide spectrum of contract sizes anywhere from $10 million or so all the way down to a couple of hundred thousand. And what we really did was, as I described, added these products and services in order to open up the rest of the enterprise kind of contingent labor wallet. Now we're now really exclusively going after kind of Fortune 200 companies on the enterprise side. So these are contracts that are tens of millions of dollars to up to hundreds of millions of dollars contracts. Now the great thing about that is it actually requires a very small sales force to go after these very large contracts. But the other thing is that the contracting period and negotiation period with these contracts are very long. So we guided to a back half of the year next year in 2026 ramp of the enterprise business. And that's a realistic time line because of just contracting times. So really, the growth in enterprise in 2026 and inherent in that 4% to 6% GSV growth rate, it's all back half weighted, which means the real ramp is in 2027. And that's one of the things that gives us such high confidence. We're actually -- we're getting tons of inbounds from our existing enterprise customers asking us to join RFPs and other things for these much larger contract sizes now even before these platforms are integrated. So we have a lot of confidence that this is a strategy that has a lot of legs.

Okay. Maybe touching on margins for a second. Margins have meaningfully improved over the last 2 years, 11% 2 years ago, now just under 29% this year. So could you talk about what's driven the improvement and whether you think the current cost structure is sustainable for where you want to take the business?

Yes, sure. Look, we're super proud of the margin journey. I think we've done a tremendous amount of optimization, both in the core business and like I said, on the enterprise side, I'll just talk first about our core marketplace business. Look, overall, this is an 80% gross margin business, right? And so -- and there's still actually ongoing optimization we can do on the cost of service line. So this is an inherently profitable business and a super capital-light business as well, right? Our CapEx is $20 million a year or something. So this is a business that is just inherently profitable. So I think Upwork, as I came into the business in 2023, had a huge growth trajectory during the pandemic as many beneficiaries did kind of invested into that. But as we looked at the business in 2023 and beyond, we really did a few things. We optimize the marketing investment. There was a lot of brand marketing that we just weren't seeing the revenue yield from. We never touched our performance marketing budget. We still have not because we get good yield out of that, and we'll continue to kind of optimize that. We focused more on acquiring these higher-quality SMB customers, but getting really good kind of ROI to CAC on that. We very much optimized the sales force, reduced it by kind of 50%. And we -- the other thing that we did with this kind of Business+ plan launching on the marketplace was we took some of the enterprise kind of lower-end enterprise product features, and we put them on the marketplace in a self-serve way so that we no longer had to have kind of account executives, more expensive kind of cost to serve in our business. We could just service these kind of smaller end enterprise customers on the marketplace. So we've really optimized the business that way. And then the third way was really narrowing our R&D portfolio to focus on these kind of 3 growth vectors that we've identified that would be game changers for our business. And so look, I've been in this kind of finance and enterprise for a really long time. And I've seen it over and over again, when you narrow your kind of investment portfolio on the R&D side and just focus it on a few really distinct things, your execution just gets faster and better internally. And I've seen that here. I've seen it in other businesses I've been in, and it's really enabled our execution to be faster while also optimizing our cost base.

Okay. That's encouraging. Maybe touching on the enterprise opportunity because that's something that is newer. Could you talk about what you're competing within the enterprise market? How are you going about the opportunity and how Upwork may be different than some of the other possible solutions that are kind of out there?

Yes. For those who aren't as familiar with, like I said, the contingent labor market, I think it's probably hard to visualize exactly why we're so differentiated, but I'll try and describe it. It's actually quite difficult for any kind of company that offers contingent labor to offer true global kind of labor arbitrage to global businesses. And that's because most of our kind of staffing competitors are actually kind of they're kind of geo verticalized, whereas we -- because we have this inherent global talent pool of 18 million talent worldwide sitting on our marketplace platform, we are able to, with now these acquisitions that we've done, offer enterprises any contract type in virtually any country in the world. And that is a truly differentiated kind of offering that no one else has done. Actually, this last quarter, when we launched the Lifted product, we announced it intentionally at the beginning of September because the largest contingent labor conference of the year was in mid-September. And they chose to highlight the Lifted product as in the keynote address, even though without even telling us, and that was because of the differentiation of what we're offering in the industry. So is super exciting. And we know from kind of the interest we're getting from our existing clients that this is definitely new and different.

So can you frame the 18 million, is that a larger pool than other companies can provide? Is it a low cost pool, it sounds like.

Yes. Look, I mean, the fact is we can offer, like I said, kind of -- because of the global nature of our talent pool. So if you think about our marketplace business today, about 2/3 of our clients are U.S. and about 1/3 international. The opposite is true of talent, right? About 2/3 of the talent working on our platform is international, about 1/3 U.S. Even greater is this pool of kind of broader 18 million who are registered active bidding on the platform, and available for work. And that even a greater proportion of that is international. So the access to labor arbitrage for these businesses is very high. We're one of the largest sources of labor arbitrage for SMB, certainly, but then also for large enterprise.

Okay. Okay. Any questions from the audience? All right. Moving on to free cash flow. So could you talk about the sustainability of free cash flow conversion in this 85% of EBITDA range that you've kind of talked about for this year?

Yes. Like I said before, so as you just articulated, our EBITDA converts to free cash flow at kind of 85% plus. So it's a high cash yield business. We're very proud of our margin trajectory. I'm also very confident we have a 35% margin target out there. We just reported 29% last quarter, 35% over the next few years. That's what, I would call it, very confident target for us because we have just inherent leverage in this business as we continue to grow. And we have continued cost optimization and things that are longer term in nature that we already have line of sight to over the next couple of years. So we feel really good about that trajectory. And as we continue to see the growth rates that we're starting to see, there's no reason that, that can't go higher over time. So there's just -- there's just a lot of kind of future cash flow growth for this business as we've been, I think, showing very well over the past few years.

Sure, sure. So there's no like working capital draw that....

So there's working capital movement in our business, right, because of just the timing. If you think about how our platform works, the timing of -- we recognize revenue when clients are billed, but then we hold in escrow some of the freelancer payments and then pay them out. So there's working capital movements from quarter-to-quarter, but over -- that just smooth out over time. And over time, like I say, our EBITDA just converts to free cash flow at 85% plus.

Okay. And there's a question...

So how do we control for, say, like for people who can't -- who aren't qualified to do what they say they're going to do or something like that? Yes. Look, I mean, we offer certain guarantees on the platform, right? And the other thing is that the marketplace dynamic, the supply and demand dynamic of the platform does also self-optimize. I mean, obviously, all marketplaces have very strict and ongoing models, making sure that there's no fraud on the platform and all that kind of stuff. But in terms of work quality, we make sure that we guarantee the quality of the work on the platform and then also freelancers gain reputation over time, just like any other marketplace business.

[Technical Difficulty]

Not software companies. So we're -- the core business is a kind of self-service, largely self-service marketplace. And the competition is really -- there are other marketplaces, although we are the -- our marketplace is unique in that we specialize in these longer-term, more complex projects. Most of the other marketplaces are kind of at the lower end of us. And then we compete with the staffing industry, but we are kind of faster, better, cheaper. Our take rates are lower, and we can just fulfill faster because there's -- by and large, there's no human in the loop trying to source and stuff like that. And then we compete with private verticalized players. So there are lots of kind of companies out there that offer just kind of one industry vertical like software development or like logo design or stuff like that. But we're really one of the only platforms out there who has hundreds of job types and companies can come and really fulfill any type of work that they want on the knowledge work side of the house.

[Technical Difficulty] Fiverr is out there.

So Fiverr, their -- that's what I was referencing, they're sort of at the lower end. So our average GSV per client is $5,000. Theirs is about $300. So we do -- we absolutely operate in the same industry, but we sort of operate in an adjacency to them.

And word on capital allocation given the strong cash flow.

Yes. Good way to end. Yes, we are very active in terms of our capital allocation strategy. Obviously, continue to do our organic investment and grow our margins. We have been acquisitive on the M&A side. We did the 2 enterprise acquisitions. We also acquired 2 kind of tech and talent acquisitions on the marketplace side. We'll continue to do that. And then we've been very active buying back stock. We're reducing our share count this year, and we'll -- we've committed to offsetting any kind of stock-based comp through stock buyback, but we'll continue to strategically reduce share count over time, particularly at these prices, I'd say.

Great. Yes. With that, I think we're out of time. Thanks, Erica, for joining us.

All right. Thanks so much, Josh. Thanks, everybody.