All right. Good afternoon, everyone. I'm Justin Patterson. I lead the internet research team at KeyBanc. I'm excited to have Ian Siegel, the CEO of ZipRecruiter, here with us today. Welcome, Ian.
Delighted to be here.
Thank you. I'm glad we just escaped the rain from outdoors with smooth sailing there.
To kick things off, for people who aren't familiar with ZipRecruiter, I'd love for you to provide just a brief overview of the business.
Sure. At ZipRecruiter, our mission is to actively connect people to their next great opportunity. We're one of the largest jobs marketplaces in the U.S. We've been around for 15 years. We have the number one rated mobile app on both Android and iOS, and have had such for almost a decade. There's over 1 million reviews in there on that app. We have 80% brand awareness on both sides of our marketplace. We span the entirety of the labor market, so we're in every geography, every job category, and we work with businesses of all sizes.
Perfect. There's going to be a lot to unpack there, but I want to start with the macro within there. You mentioned the labor market. That was certainly an atypical labor market even before AI came into play. I'd love to hear just what you see within the labor market today and how you're helping companies really navigate the AI transition.
Yeah. Those are two very big questions. First off, what are we seeing in the labor market? Post-COVID, there was a run-up in the labor market that was historic. Then for the last four years, the labor market has effectively been into a persistent decline. Hiring in America has just been falling until 2026, when we came into our first period of extended stability. In 2026, I would call the labor market stable but subdued.
The reason I use the word subdued is because if you look at every third-party data point, but most of us will go look at the BLS data that's available, you will see that both hires, the absolute number of people who are getting hired in the U.S., and quits, which is the thing that drives hires more than anything else, they're both hovering around the same low levels they were at all the way back in 2015. So while the decline has stabilized, we have not yet returned to either what I would call a normal, healthy labor market, or seen the macro start to improve.
Got it. I think one of the interesting things around AI right now is this whole notion that we're going to vibe cut away some jobs, we're going to have massive layoffs across the board. Yet recently, you published a report that's showing some opposite signals in there. I'd love for you to unpack just what you're actually seeing from the labor demand, even in a subdued environment.
Yeah. We study this question from a number of different angles. But in reference to the survey you're talking about, we recently completed a survey of over 10,000 businesses, and roughly 70% of them were SMBs, and the other 30% were in the enterprise category, or they had 5,000 or more employees, so they were of a larger size. What we discovered was really interesting. So across all those businesses, over 90% of them said they are already using AI in some form or fashion inside of their business. So AI is permeating every size business in America.
But the biggest surprise, at least for me, was that 35% of those businesses say, as a direct result of using AI, they are actually planning to increase hiring and/or have already begun to increase hiring, which is counter to the narrative that AI is coming to destroy the labor market. There are a bunch of other signals you can look at that support this premise that so far, AI is not having a significant negative impact on the labor market. The one I find most interesting, and I will just call this one out, is everyone has predicted that engineering jobs were going to be the first to go, and they were the most at risk from AI disruption. It was the thing that all of these large language model companies were focused on in terms of skills that they have hyper-optimized.
If you just go back to the beginning of when Claude Code was released till now, engineering hiring is now up 10% from that period. So even though we have all marveled at what Claude Code is capable of, it has not decimated, at least, the labor market for engineering.
If anything, it actually seems like it has made the software engineer more accessible to the small business who might have been locked out of paying for that person before.
I get paid to hype the labor market, so when I say things like this, people are like, "Where is your data?" What I would say is that if you go back in history and you look at every major technological breakthrough, it does not matter if you are talking about ATMs and what that did to bank tellers, which was to massively increase them, or if you talk about spreadsheet software and what that did to accountants, which once again massively increased the number of them. Every time you make something easier to do in our society, history tells us we will want to do more of it, not less of it. So yeah, it will change the nature of how those jobs are done. Bank tellers do not do normal standard transactions anymore. They basically upsell you to higher value products.
The same thing is happening in engineering and across a wide variety of other categories where AI is making things so much easier to do that more and more businesses find it accessible to do those things.
Exactly. It's an abstraction problem. The nature of work changes, but we still have people running around at the end of the day.
Yeah.
Makes a lot of sense. Shifting gears toward just your financials and kind of bridging that macro side, I thought this last quarter was really interesting, where I believe that your end market was perhaps growing flattish, call it, yet you grew 5%. What are you attributing to the drivers for your outperformance?
Yeah, last quarter was great for ZipRecruiter because in spite of this subdued but stable labor market, ZipRecruiter grew top-line revenue 5% year-over-year. That contrasts with Q2 a year ago when we fell 2% year-over-year. So there was a seven-point swing between the two. What is very clear to us is that it was our product improvements that directly led to the increased growth in our business. We have spent a considerable amount of time and effort focused on one thing, which is driving conversations between employers and job seekers. This is not about increasing the number of clicks. This is not about increasing the number of applies. This is specifically about getting a human to talk to a human.
You may say, "Why did you do that?" We have long seen a correlation in our business between humans talking to humans and increased everything, revenue, engagement, applies, ongoing usage, reactivations. We have consciously been trying to drive up that engagement rate. What you saw in Q2 was the stack of improvements that we have been launching the back half of last year and this year really coming together, and I will just call out one of them in particular. We foundationally launched this brand-new search engine that was retrained to focus on delivering not just candidates, but quality candidates to employers. We were excited because it increased the percentage of quality candidates that got delivered in a step function style fashion by 34%, which is incredibly rare for something new like this to do.
You may say, "What does that mean?" I say, "That's a fair question." The real thing that you want to look at is if that's true, then you would expect the engagement rate, where employers who receive candidates, you would expect to reach out to those candidates at a higher frequency because they're getting more qualified candidates. That's exactly what you saw in Q2. The engagement rate through both the new search engine and a variety of other improvements has doubled year-over-year. So when candidates apply, they are twice as likely to get a response from the employer, and that's magic for a marketplace like ours.
Got it. I want to touch on a point you said there in that it's rare to see that type of result from a new product like that. When you look at the product roadmap ahead and just the enhancements you can make on the search model, how do you think that changes the business over time?
We're seeing it play out here where, if you go all the way back to the S1 ZipRecruiter put out when we went public, we were talking about AI and being an AI-driven marketplace before there were LLMs. Because back then, we were really focused on machine learning and meta learning and deep learning. With each new technique and tool that has come out, we have layered it into what we do to increase the rate of conversations. What LLMs are very good at is not actually the matching component of what we do, but rather the assessment component.
It takes the treasure trove of data we've built up over 15 years, the billions of interactions between employers and job seekers to train algorithms to use both direct matching plus the wisdom of the crowd to create the pool of candidates that the algorithms believe would be qualified for a job. But what LLMs have let us do is then put a layer of observation against that pool to assess it the way a human would of who would be most liked by the employer. And that's just one example of that assessment because it actually goes both ways. But that has been part of the secret sauce in massively increasing the rate at which qualified candidates apply and response rates from employers.
In every way, LLMs and their capabilities have been making ZipRecruiter better, and I certainly think that the future is not just going to be about matching, but it's going to be about driving that engagement between the two sides.
Got it. And you mentioned consumer awareness in there. Obviously, for helping more people find jobs in a tough environment, there's probably going to be some word-of-mouth benefits from that. I know it's early with this new product, but are you starting to see more changes on the traffic composition, what's coming in organically versus paid from this new product?
I think ZipRecruiter has tremendous brand awareness on both sides of our marketplace, both with employers and with job seekers. It bounces around 80% for both. And that is operating leverage for our business. We don't need to keep paying to bring large volumes of people to both sides. We are an answer to the question both sides ask. Where do I go to look for work? Where do I go to post a job without them needing to go through a third-party service? However, a large portion of the population still does go through something like a Google, or now, interestingly, LLMs, whether it be Claude or whether it be ChatGPT. We did a full integration with ChatGPT in Q1. We did an integration with Claude in Q2, plus an upgrade to the original ChatGPT integration we did.
The traffic that is coming from there is still de minimis as a percentage of overall traffic, but it is growing rapidly, probably the fastest-growing source of traffic we have. The users are incredibly high on engagement scores, meaning they are active job seekers. They are applying to jobs. They are currently looking for work. Which is what you would expect, but why it is so exciting is, when you look at sort of the competitive landscape, this is a brand-new channel that we are at just as it is emerging, and we are immediately integrating and one of the initial launch partners for both of these new channels. Which means we are on equal footing with every would-be competitor, and we are harvesting our fair share and more from these sources. So they are turning into a competitive advantage for us.
Got it. Perhaps to pull that thread just a little bit further within there. When we talk to Google, it is very clear that LLMs, and even just what has been in AI Overviews, AI Mode, is driving just much richer conversations. You are getting searches that did not exist previously.
It sounds like data is one of the many inputs into your business. When you look at kind of that initial sample in there, how does that corpus of information change your ability to match down the road? How does even having these longer, more conversational interfaces change what you might put in front of job seekers down the road?
Yeah. I think of what the LLMs are doing, not purely from the vantage point of how the search engines are changing and incorporating it into their interfaces. Because fundamentally, they are just helping you skip a step there. They are saying, "You used to research, what does a plumber get paid?
they'd show you a link, you'd click on the link, then you'd go to a page that told you the answer. Now they're just saying, "I'm going to put the answer directly in front of you." What's more interesting to me about these LLMs are the capabilities they unlock to make things easier for businesses who historically have had to pay for either a person internally or an outside party to perform a task for them. The most basic example in recruiting is leveraging a resume database to identify candidates and then do outreach to those candidates. Manually is a lot of work. A lot of people pay a third-party recruiter to go do that work for them. In Q2, ZipRecruiter launched what's called Smart Outreach on our resume database.
Now one person can go identify the type of person they're looking for, and the software is very smart and can say, "Okay, I found a lot of people who look like that." More importantly, it can run an outreach campaign to those people so that that person doesn't have to be like a master of organization who's like, "Okay, I sent my first message this date. Did I ever get a response? No. I'm going to do a follow-up on this next date." That whole campaign of follow-through happens programmatically through the LLM. Sounds incredibly human. It's outreach from an employer to a job seeker, which is obviously job seekers' far preferred methodology of job seeking, as opposed to going and proactively searching on their own.
It's leading to some concrete outcomes and is part of the overall improvement in response rate between employers and job seekers that we're seeing. Those kind of capabilities are where the superpower of LLMs lie.
Got it. That sounds like something that could open up future pricing potential because you're saving the employer a lot of time spent in there. If the models are working as they should, you're delivering higher quality, more relevant matches, which means they fill that seat higher and can start getting a productive worker. When I think about recruiter fees, those are very, very expensive in there. So when you look at your pricing today-
versus the broader industry, what type of pricing uplift could you see down the road?
There is substantial headroom for pricing increases inside of what we are doing at ZipRecruiter. I think unquestionably, that is a lever we will pull at some point. However, when you look at this current backdrop, we are in a market where hiring in America was in decline for 3.5 years and has only now started to stabilize. In that environment, we are growing our top line. We are growing our customer count, which is extraordinary, and I think something that we plan to keep investing into from the standpoint of taking that market share, increasing the scope of our customer base. We will defer maximizing pricing until we start to see a market where we are not able to make those gains and/or the macro starts to actually move in a positive direction.
For now, though, I think that the pricing lever is very clearly available to us, in particular, when you look at offline recruiters and what they are still charging, in the majority of cases, in order to secure you talent. I am just highly confident that that will still be there as the market either continues to stabilize and/or even reverts to what we would expect as a more normal, healthy rate of hiring in America.
Makes sense. That seems like a natural segue to just talk about the marketplace growth itself-
Yeah
in there. You have large enterprises on one side, small businesses on another area. Could you expand on just your go-to-market approaches toward both sets of customers?
Yep. ZipRecruiter started in the SMB category and then has moved upmarket into enterprise. When we first started, we were 100% SMB, 0% enterprise. Today, we are 76% SMB, 24% enterprise. We run very different playbooks against those two segments. They require different strategies. We do a lot of marketing through a variety of channels to bring prospects to the website, where they largely self-convert, but we have an inside sales team that works against those leads. We also have a large outside sales team that goes and sells to those enterprises and tries to get them to come in and use ZipRecruiter and is having a lot of success doing that, particularly with the advancements that we've been making in our products.
Now, as you look at that mix over time, I think you said roughly 75/25 right now.
Yep.
I believe you've got a 50/50 target out there.
Yep.
What do you see as really the key steps to driving enterprise up to 50%?
Enterprise is a complicated animal relative to SMBs in that most enterprise companies are using a third-party software called an applicant tracking system in order to be their canonical source of the jobs that they have open and they want to operate out of there. If you want to work with enterprises, you have to do integrations with as many of these companies as you can. We've been doing these integrations for over a decade. We have something north of 180 ATS integrations in place. On top of that, these are sophisticated buyers that are used to a pay-per-click, Google-style performance marketing methodology in terms of buying. There's a bunch of stuff you do there around campaign optimization or bid optimization that you have to layer in there.
You need an outside sales force to go work with these accounts the way they're used to being worked with. All of that groundwork has been laid at ZipRecruiter. We have a lot of momentum with the enterprise side of our business. Both sides of our business, however, are growing. In this period of stable but subdued labor market activity, every part of ZipRecruiter has started to show benefit from the product improvements that we have been making. That trajectory to a 50/50 split, which is reflective of how hiring works in America, half of hiring is done by enterprises, half is done by SMBs, is continuing apace and is following its expected course.
Makes sense. I think another nice tailwind you've had, beyond just the product innovation, is industry consolidation, where a lot of some of the companies have been taken private or taken over by a strategic seem to be facing some challenges right now.
Could you talk through how you think that benefits market share going forward? Have you seen the bulk of that market share lift already, or do you think there is more to come down the road?
There has been a very tough four-year period for the recruiting industry, and a lot of change has happened over that period. If you look at the landscape, there is a large, disaggregated, high volume of job sites out there that work a wide variety of verticals and specific niches. They tend to be on the smaller side. There is a handful of larger players, some of whose names you guys are familiar with in this room. All of them have suffered over the last four years, and a shakeout has occurred where a significant amount of market share has moved from the smaller guys to the handful of larger players. Amongst the handful of larger players, there has also been a dog fight for market share. It is not limited to with each other.
It has also very much been with where the bulk of the dollars in our category lie, which is the offline recruiting world. The TAM on recruiting in the U.S. in particular is split between really three groups. You have the in-house recruiters who are paid by large companies in order to do their recruiting for them. You have the contract recruiters that are outsourced to and are usually paid something in the range of 25% of first-year salary. Then you have the temp labor market, where the recruiting firm acts as the employer of record and sends talent to a business. All of those have suffered, along with the online companies, during the downturn in recruiting. However, market share gains have been made as technology advances continue to improve.
You go back to that smart sourcing product that we recently launched against our resume database, and you can see how the fundamental function of a recruiter has been largely replaced by software. That is creating a real opportunity for disruption against where the bulk of the dollars go and market share grabs.
Got it. I would love to hear how you are thinking about allocating capital in this environment. You just paid down debt opportunistically, realized a lot of value for shareholders. As you look at a business that is free cash flow positive, intrinsically high gross margin, what do you see as the natural uses of incremental investment right now?
Yeah. Just to explain what you referenced, we recently paid, we had about a $500 million of debt that was in the form of a bond, and that bond was trading at a substantial discount. We made the decision to go try and buy back a large portion of that debt and harvest that discount. Delighted we were successful. We took $295 million of the debt and retired it. We got a $65 million discount on that repurchase, as well as we will get north of $50 million of interest savings over the next 3.5 years. It was a win by every measure. It is a unique opportunity. We took advantage of it. We still have over $170 million on our balance sheet, more than enough to fully fund every one of our strategic objectives that we have.
When I look at the use of capital and what we have used it for historically, our philosophy remains the same. It is organic growth first. What can we invest into that will accelerate the growth of our business? Inorganic growth, so opportunities to either make acquisitions or to make one-off single moves that would propel the business top-line growth and/or improve the bottom-line margin. As a final use of capital, we would look at either defraying our debt further or potentially continuing to repurchase our stock as we see it as being undervalued.
Got it. I think we have time for one more question here. If I step back, a few years back, the business had peaked at about $900 million in revenue or so.
You're about half of that today. As you think through just the factors to get back to that prior high water mark, how do you think about the role of both the current macro, call it remaining stable-ish, and what you can control on the product side to getting back to that level?
Well, I feel really confident that we are just at the beginning of the impact that the product improvements can have on our business. As a result of that, in a stable, ongoing macro at this depressed level, we can continue to grow. I also feel confident that if the macro were to continue to be the cycle it has always been, and it were to actually start to improve as it has always done, that our rate of growth would accelerate and that it would be much faster to return to the historic highs that we have previously reached and the 30% margin target that we continue to inform our investors is where we believe we will ultimately wind up.
Great to hear. With that, we are out of time. Ian, thank you so much for joining us here today.
Thank you.