I think in the interest of time, we're going to get going. I think we're also going to close some of the doors in the back, but people are going to be no doubt milling about. It's great to have ZipRecruiter back at the conference again this year. Ian, thanks so much for taking the time and giving me the opportunity to have the conversation. You've been a very consistent guest for us over the years, and we always appreciate that.
Yeah. Thanks for having us again.
Always good to talk, and I always do like to start with a little bit of, for those who are either listening or in the room and don't know the story as well, you've been on a bit of a journey. I also think one of the most interesting things are you're a company that was talking about AI before anybody was talking about AI, which I always like to give you some credit for.
T hank you.
I think that's worthy of some credit because talking about AI is certainly in vogue as opposed to forward-looking. Talk about the journey the company's been on. You've been through elements of macro cycles, product cycles, platform changes. Bring us up to speed on the journey you've been on.
Yeah, it's like telling the whole story feels like it's getting so long. I feel like I'm going to frame it just in the most recent period. After launching this startup at my kitchen table and after many years of product evolutions where we ultimately got to using AI for the first time ever in the job category, I think we were sort of on the bleeding edge of doing that. The company experienced tremendous success, grew fantastically well. We were delivering quality applicants who turned into hires at record rates for our industry. We were able to ride that momentum into taking the company public in 2021. Everything was going fantastically well until the last 3.5 years when a macro downturn hit the U.S. economy. Hiring effectively went backwards every month for 3.5 years. That's bad if you're a recruiting business.
Fortunately, hiring has stabilized in 2026, albeit at levels we haven't seen for over a decade. In spite of that, the good news is ZipRecruiter is back on a growth trajectory that's coming from product improvements that we have deployed and that I'm sure you're going to ask me a question that gives me an opportunity to talk about.
I will. Let's start with that macro trend first, because obviously what you're seeing in the last most recent periods has been a nice change of pace relative to, as you talked about the last couple of years. How are you thinking about what the drivers could potentially be for a healthier end demand side of the equation for the labor market more broadly? What are you watching for to get increased confidence signal in an upward trend?
This is a great question because we regularly survey both sides of our marketplace, and in the most recent survey where we talked to, I think it was 10,000 businesses, and that ranged in size from very large enterprises to SMBs. What we learned was a bunch of things. We learned that AI is not a boogeyman coming for the labor market, that AI is more likely to be something that creates job opportunities than removes it. Happy to dive into that. More importantly, what we learned is businesses are doing well. Businesses are confident. Many of them are experiencing either record levels of top-line performance or record levels of profitability, which is normally a recipe for businesses to expand and increase their hiring, but they're not doing it.
The reason they're not doing it, and the overwhelming message that we're getting from our customer base is they have anxiety about the economy. They are worried about investing into an economy that may be soft sand underneath them. They're in a wait-and-see mode, which is part of the reason why they're experiencing these record profits, because they're not keeping the investment up with what you would consider the normal healthy rate in order to support the amount of business demand that they're experiencing.
W hen does that reach a friction point in your mind where you could only stretch that equation so far in certain pockets of the economy, where you can only extract so much output per unit of labor that you see today as opposed to what it might look like six or 12 months from now?
I'm not a full-time economist, so I may not be fully qualified.
I kicked off this conference by interviewing our economist.
Let's get him in here, yeah.
He did sound that there is a lot of productivity gains that have come out of the labor force, and that typically is not a long-duration narrative that you can underwrite.
There are a combination of elements that are contributing to the very strong likelihood that we are going to either currently be at the bottom of the cycle or very soon we will be. Businesses' performance is very strong. There is a shortage of skilled talent, in part because of people hugging the job that they have for fear of difficulty finding a new job. Also, in part because of immigration policy, where it's become much more difficult to bring talent in from other countries. Then there is this generational dynamic going on where every year the size of the population now that is moving out of the labor force, that is aging out of the workforce, is at the same size or smaller or larger than the incoming inbound youth who are matriculating and entering the labor force.
All of this is creating pressure for businesses where there is a need to hire and the amount of available talent has shrunk. In a normal cycle, what you would expect to see at these moments is that you would start to see an increase in hiring and an increase in willingness to pay. I think over the next six months, it should be very telling to see whether or not we have actually experienced some sort of fundamental structural change in the economy, and this was the last cycle, and we reached the bottom of it, and it is just going to be the new normal, or if, in fact, the cycle is a cycle, and it is not going to start to climb again, which I would argue, based on history, is the much more likely scenario.
Okay. The other thing you said there that I think has become a little bit of a theme of this conference relative to when I have been in this pocket of the world three, six, nine months ago, which is the fear of a jobs apocalypse seems to be moderating. I think that has become, across a number of conversations I have had, a theme that has been interesting to hear from leaders of organizations. How has your view evolved, if it has even evolved, or maybe you were there to begin with, about how AI and the labor market might have a correlative effect on each other in the years ahead?
Well, certainly, and I like to go to the data in questions like this. There has been a lot of anxiety about whether AI is coming to eradicate large swaths of the jobs that currently are at play in the labor market. From what we can see, again, both from surveying our customers and from third-party data sources, our surveys show that of the thousands of businesses that we talk to in them, the majority of them, the vast majority, over 90% of them, have already adopted AI to some form or fashion. For them, AI is proving to be a productivity boost, and as a result of that productivity boost, they intend to increase their hiring, at least 35% of them do.
When you look at the general data from third-party sources like the BLS, there was a lot of concern in particular about a couple of key job categories being things like customer service and software developers, being jobs that would be the first to be eliminated. Let us just take engineers as a great example. At the end of last year, they were down roughly, the postings for engineers were down about 10%, and if you go to this year and this quarter, they are up 7%. A lot of what the anxiety was creating in the labor market was a mirage. It was not real.
We are in fact seeing the same pattern play out right now that we have seen play out in every major technological breakthrough that made something easier to do, which is when something gets easier to do, the demand for it increases. Right now, engineers as well as customer service reps are still jobs that are in existence, and in fact, demand is rising.
Okay, interesting. Let's turn more to your business specifically. In terms of the trends you're seeing, how would you characterize some of the dynamics of the way in which smaller and medium-sized businesses are interacting with your platform as opposed to larger enterprises are interacting with the company from a demand perspective?
Every part of our business is growing now, but I've got to put a very important piece of context on that, which is this is not an endemic reality of the macro that is causing that growth. This is the direct impact of product improvements that we've been deploying over the last nine months, but in particular over the last six months, where the compounding effect of those improvements is leading to satisfaction and increased investment amongst our customer base, and that is what is driving the growth of those.
It is very clear from looking at this sort of organic macro data, yes, we have seen stabilization of the macro. Instead of a three and a half year decline, we're starting to see things flatten out. T he truth of that flattening out is it's flattened out at a level that is depressed so significantly. We haven't seen levels this low since 2015. It's been a decade since we've been at where we're at. Businesses have not yet, of any size, in a material fashion, not SMBs or enterprise in any sort of distinct breakaway fashion, started to increase their hiring again.
Okay. When you think about targeting the larger enterprises, though, what kind of building blocks are you trying to put in place to capture more share in that part of the market, and how should we be thinking about that as a market opportunity over the medium to long term?
There are two fundamental strategies at ZipRecruiter that are driving our revenue growth. One of them is we have an intense focus on driving the two sides of our marketplace to have a conversation. A conversation can be anything where an applicant applies to a job and then the employer responds to that application with some sort of message. It can be a scheduled pre-screen call. It can be a scheduled interview. There is a variety of things that qualify as a human talking to a human. That is fundamentally our roadmap. That is the most important strategy we execute against, and almost every product we have delivered over the last year has been explicitly focused on that. That is what has been driving a lot of our current results. That is strategy number one.
Strategy number two is we have been pushing for a mix shift in our customer base from being where we started, we were 100% SMB, and then we moved into the enterprise market. We are now at 76% SMB, 24% enterprise. That contrasts with the same time last year where we were 78% SMB and 22% enterprise. T here has been a steady march of increased penetration into the enterprise market. The U.S. economy has a 50/50 split between the two sides of the labor market between these two companies of different sizes, company classes of different sizes, and that is going to be a growth lever for many years for ZipRecruiter.
Okay, understood. You referenced a little bit earlier how AI can impact the landscape, but I want to go a little bit deeper on how it is impacting the company. Maybe just a couple going down this road. First, the deployment of AI inside your company, how is it changing the pace and cadence of product development, cost efficiencies, way in which you go to market? Talk a little bit about the way it works inside the company before we get into some of the elements of the broader landscape or user-facing AI.
I think ZipRecruiter is not unique in saying this, but AI has been an extraordinary boon to the productivity of ZipRecruiter. The quality and velocity of new features going out on our platform in the last nine months is probably greater than the sum total of the previous four years, and we have been able to do that while simultaneously reducing our overall investment in R&D as a percentage of total. We're getting incredible operating leverage out of AI as just a means by which to get our ideas into the product. Further, AI has permeated the product. We're not big believers in the chatbot interface that partners with a user or a customer to try to help them do things like write better job descriptions or write better resumes. Anything that slows either side down doesn't tend to be a winner for us.
But we are leveraging AI at multiple points in order to do things either with our algorithmic matching or what we call socially engineer the two sides to engage more quickly. A number of product features have gone out that capture this and epitomize what we are thinking of as our product philosophy. One of which is the new search platform we just brought out in Q2 has been transformational. Instead of optimizing and training our algorithms to go seek more clicks or more applications, because we're a closed-loop system, we had the data to train them on actual employer response to applications, quality applications, if you will. Using that as the training data set has increased the number of qualified candidates by 34% and increased the response rates by employers to applications by double. Doubling the amount of conversations happening on our platform is transformational.
Just one more example is we launched a product called Smart Outreach. This is just a mechanism to allow employers to better utilize our resume database, which has over 50 million people in it. They can basically create the job that they're interested in, and then AI will go do all of the sourcing for them, where it will reach out to the appropriate matches. It will have a cadence of messages to try and induce them to become an inbound warm lead for that employer, and satisfaction with this product has been tremendous. Those are the ways in which we're using AI. It's not as a, I don't know, Microsoft paperclip that runs along beside you and gives you advice along the way.
Something you're alluding to there that I want to go a little bit deeper on, which is traditional job queries were very search-oriented in nature. Is there a way in which you're thinking about, it sounds like it is, reordering the funnel to make it either more conversational in nature or to generally change the discovery process on both sides of your market?
I would say we're certainly open to that, and we're deeply thoughtful and experimental with that. It's very interesting. The bulk of the users on ZipRecruiter are what we call active job seekers. These are people who are saying, "I need to find a job. I'm actively engaged in search." Interestingly, the single fastest-growing and new source of traffic for ZipRecruiter for job seekers is LLMs. Whether it's Claude or whether it's ChatGPT, we see an influx of job seeker traffic from these two sources. Not only do we see that influx, and not only is it growing rapidly, but these are the most engaged job seekers we've ever received.
It totally makes sense because the conversation is happening on those sites where someone says, "I'm leaving the military and I'm trying to enter civilian life. How do my skills translate? What jobs should I be looking at?" The AIs are fantastic at answering questions like that. Once they know the type of job the person is interested in, though, they still send the query to us to get the set of jobs to show the individual. Once the individual clicks on one of those jobs, they are then sent to our site. I think conversational UI is something that makes sense for LLMs. It's the expectation there. I think on our particular site, jury is still out whether we want to move to conversational UI or not.
One thing that's been coming up at the conference over the last couple of days has been the balance that companies are trying to strike in traffic that comes to them from these agents relative to direct traffic and trying to make, once someone's landed on the site, more AI infused and less friction. How do you think about making sure you don't lose control of lead generation and you become more overbalanced to traffic that comes from LLM agents over time?
First off, it's been 1.5 decade of investment in building the brand for ZipRecruiter. We're at 80% brand awareness. We're the answer to the question that when people say in their head, "Where should I look for work?" We're one of the top ideas that immediately pops in their head, circumventing the need for them to go to Google or an LLM. They know they can find jobs on our site. That's number one. Number two, what makes ZipRecruiter special fundamentally is not any specific feature. It's not any specific algorithm. It's the data that we've collected over those 15 years.
All the billions of interactions between job seekers and employer jobs, and then employers' reaction to those applications is what allows us to do the training on algorithms that creates this magical experience where seemingly the right jobs for you are put in front of you, and when you apply to those jobs, you have high probability of getting a response or engagement from the employer. That data does not exist. That data cannot be taken by an LLM. LLMs have recognized that and partnered with us. We have Integrations with both Claude and OpenAI, and we were there at the beginning, and it's a really exciting opportunity for us. We have no fear of those sites taking our business away from us.
Okay, understood. You've also done a number. Well, at least one in particular that I wanted to focus on, but when you've done acquisitions in the past, I'm always curious about how they change your go-to-market strategy or what you're offering to folks in your marketplace. Breakroom comes to mind in terms of the way we think about it, but talk to us a little bit about how you continue to make decisions about what you're going to build, what you might acquire, how it might speed up your time to market, how it can change your go-to-market strategy, your product strategy. Just go a little bit deeper on that as a topic.
O ur product roadmap of driving conversations is derived from a very simple and intuitive insight, which is the more conversations an employer has with job seekers on our site, it's linear, the more they pay us. The more conversations a job seeker has with employers, the inverse of that, the longer they stay active in their job search and the more jobs they apply to. This is why we decided we were going to target everything to this and try to make ourselves a marketplace of conversations. The way I look at acquisitions and the performance of features that we have deployed is entirely around how much do they increase the rate of conversations. You look at something like Be Seen First, which is a feature we launched at the beginning of the year.
This is a product that lets a job seeker say, "I'm not only applying to your job, I'm going to raise my hand and show you how enthusiastic I am by telling you why I'm qualified and why I was excited to apply." That message being included with their application doubles the probability that they're going to get a response from the employer, which makes sense. Employers are beleaguered with resumes that look perfect because AI is very good at helping people write resumes that look perfect. How do you differentiate between these candidates? One of the ways is by letting the human speak with their own voice.
Another product we have is Breakroom, and in a similar lens, Breakroom allows customers to, on their job listings, not only show the ratings and reviews. For those that don't know, Breakroom is kind of like Glassdoor, but for frontline workers. Instead of having ratings and reviews of employers from those workers about their bosses, it just lists a bunch of facts about what it's like. How long do you have to be on your feet? How long do you have your lunch breaks? What is their pay schedule? Things that people who work frontline worker jobs really care about. These jobs represent over 60% of the jobs in the economy.
What's so interesting is when you buy Breakroom on our site, it lets you put onto your job postings information about your business that also has this third party, like people who've worked there before, data inside of it. That tremendously increases engagement with jobs. It's the same thing, where now the company is basically enabled to use their real voice to talk to customers and not their traditional job posting voice. In a way, it's almost like getting a pitch from the company, and that drives tremendous interest from those job seekers.
Okay, understood. Wanted to double-click on competition. As the online recruiting and overall hiring and recruiting landscape continues to evolve, how do you think about your current competitive positioning in that landscape? What are the most potential for building moat around competitive advantage as you see playing out when you look against who you compete against today?
I think we are the only company that has taken the particular strategy of not optimizing for clicks, not optimizing for applies, but actually optimizing for one-to-one conversations between the two sides of our marketplace. That's because of, again, the strong correlation we saw with both the amount of money employers would pay us and the length and duration of engagement that we would get from job seekers. It is working. If you look at us a year ago, our business top line went backwards 5% and our margins were 9%. If you look at the run rate that we're on this year in the scenario we've outlined, we're back to positive single-digit growth, and our margins are increasing at the same time to between 12% and 14%, is the range we've given.
That is being entirely driven by this product strategy, where we are fundamentally increasing the foundational satisfaction with our service on both sides. That creates enduring growth. That's not a price increase where you get a one-time, one-year bump. That's a compounding benefit over time. We're really excited about the strategy, and I would say the biggest difference between us and our competition right now is ZipRecruiter is very much in a market share volume growth mode, and I think our competition is more in a yield optimization mode, where they're trying to figure out what the limits are on how much they can charge for their products. I think this is like a moment where we can take a lot of market share as a result.
Okay. Maybe building on that, you talked about what happened with growth and margins looking backwards, how you've guided going forwards. I think during a period where you faced headwinds, you took a lot of proactive moves with respect to the costs you could control and some of the incremental margins in the business. If the business returns to more steady state growth, what's the messaging for investors about how much you'd want to invest behind signals of growth, how much you might build on your last answer about potentially being a share taker, and how investors should think about the balance you might try to strike between capitalizing on a growth environment and still delivering on margin trajectories?
Well, I think a lot of evidence exists for investors of our philosophy about how we run the business. Since we've gone public, in spite of there being a 3.5-year period where the entire labor market was hiring less every month for those 3.5 years, we've never had a year where we weren't cash flow positive. Further, you can also look at what we did with our bond debt earlier this year. We had $550 million of bond debt. We were able to retire over half of that for a $65 million discount. We saved ourselves $50+ million in interest payments over the next 3.5 years. We're very thoughtful about the capital that we have available and how we run our business. If you look at the growth that we're getting, it's product-generated growth.
As I said, it's going to be enduring and compounding. That's our belief about the way that we're driving this growth. It doesn't require us to significantly change our cost structures in order to harvest and take advantage of that growth. That said, we've been marketing across a multitude of channels for over a decade. We have highly sophisticated early signal when the markets are either increasing in demand or decreasing in demand, and we're able to adjust our spend rapidly. We will take advantage of opportunities where demand is rising, but that should translate into increased top line that is observable because the efficiency will be getting better in those scenarios.
Okay, perfect. You did talk about the decision you made on the debt repurchase. With that as a signal, how should investors think about what the right capital structure is for the business medium to long term, and how that might feed into the scope to also continue to return capital to shareholders as you reach a capital structure that you feel is where you want to operate in more normalized environments?
I think our philosophy on what we think is the balance of where we want to invest our dollars has stayed consistent throughout, which is organic growth is always going to be our top priority. Our secondary priority will be inorganic growth. Obviously we contemplate M&A. We have downside. I think we've moved from $400 million to $170 million of cash on hand, but that's a pretty sizable war chest, so M&A is still on the table. It would be return of capital to investors, which can happen either through the retirement of additional bond debt or through share repurchase. That is a tertiary priority. We've been highly opportunistic, and I would argue, very strategic about how we have tackled that capital return. Our priority certainly remains organic growth, and we continue to look for opportunities there.
Okay. We only got a few minutes left, but when you take a step back and if we're lucky enough, or I'm lucky enough to have this conversation with you in a year's time, what are the biggest strategic priorities you're focused on? How are you aligning growth investments to accomplish those priorities? What's your sort of 12-18 months to-do list?
N umber one, we've just gotten started increasing conversations, and what's been so fantastic about it is these were not empty calories. Every tactic we have pursued here, where we've moved that metric, revenue has continued to move with it, so it feels great to have a product-led growth strategy. Number two, we need to penetrate enterprise. We know this. We want to move our marketplace to a 50/50 split, which is reflective of the U.S. economy. The opportunity there is sizable. Enterprises have persistent hiring needs and deep pockets. It's a community that we know we need to be working with more closely.
Okay. I look forward to hopefully having the opportunity to check in on those, not only on the earnings, but in a year's time in person. Thanks so much for being part of the conference. Please join me in thanking ZipRecruiter.