SKLTFSKLTFOTC
Loading
Operator: Ladies and gentlemen, thank you for standing by, and welcome to the SEEK Limited Half Year Results Call for 2026. [Operator Instructions]. I will now hand the call over to SEEK Limited CEO, Ian Narev. Please go ahead.
Ian Narev: Thank you. Good morning, all. We are here on the lands of the Wurundjeri Woiwurrung people of the Kulin Nation. I'd like to pay my respects to the elders past, present and emerging. We've got the usual crew here, Kendra, Peter, Simon, Grant, Dan and Pat. And thank you all very much for joining us. I'd only to tell you that these results happen at a pretty volatile time in terms of equity markets and today is not a sales job. The numbers are going to speak for themselves. And we'll add to that our perspective on the main drivers of the numbers and what that means for the future. In terms of our take, I've been generally encouraged by our Board to overcome my natural conservatism and talk about the business. So let me say that over the next hour and in the course of discussions we'll have with many of you over the coming days, I think it will be clear that as a management team, we have a very strong sense of confidence in the position of the business today and in its prospects for the future. We've got a combination of market leadership positions, brand strength, customer trust, technology backbone, product depth and breadth and perhaps most importantly, the data that come from all of those strengths. And that puts us in a unique position to continue and accelerate years of innovation. That takes a lot of hard work, but we've got as good a starting point as we could want. And the momentum, as I said, is clear from the results. As you can see from Slide 6, shares up, yields up, operating leverage is clear. We've got a record dividend. The balance sheet is strong. We know that the translation to the bottom line still needs to come a bit more. And particularly, we can answer questions about the Zhaopin write down in Q&A if people want to know more, but we are well on the way. This is the latest in a series of results we've delivered exactly what we said we would and probably a bit more. This is the best of that series of results. And as we discussed at our Board meeting yesterday, it's probably as good a set of operational results as SEEK has ever delivered. So the momentum is accelerating. I won't go into the detail on Slide 7 and 9, but the message there is very clear. Short-term results are as planned. Progress against our core strategic priorities is very strong. More important than that are the drivers of those results, and I'll just spend a minute on Page 10 because those drivers of the result are the connection between the strength of the results and the confidence in the future. On the top box, put very simply, we are 1 decade on from the establishment of our AI function with that name. We're 2 years on from elevating the function to my direct reporting line, and we can see the benefits of years of doing that investment right through the results. And you can see hundreds of product releases in the last 6 months using AI capability, placement share and yield growth, not prospects, but delivered these results. A high proportion of the benefits come from the data advantage we have. And a very important part of, again, not what we think, but what we've learned is that in this environment, and we believe in the period ahead, competitive advantage doesn't come from AI models. It comes from the extent to which the models can access unique proprietary data that other models can't, and that translates into tangible benefits for customers who, as a result, choose you and will pay for what you offer. You understand SEEK well enough to know we run a marketplace with very complex preferences, not on one side, but on 2 sides of the market. So using data to help elicit those preferences and more significantly to match them creates the value that, as I said, makes people choose us and pay, and that's not a thesis. The evidence is right through the results. It's evident in placement share, it's evident in yield growth, and we've got a long way to go. And you can see that in the second box there, we talk about having roughly 0.75 billion data points, most of which can't be replicated and scraped. And we know from daily experience how to translate that into the products that matter to customers. The last thing I'll say before handing to Grant is that the matching that we're seeing is better than ever, and we really feel we are still only at the start of what we can do. And I just want to dwell on one point, particular words in that first bullet point where we say the most obvious benefits from proprietary data, not just what customers want, but how realistic is it. And I think there's one thought to keep in your head here. There are many models, interfaces that can do a good job of having conversations with people to decide what they want. The big question is what do you need to show people how realistic that is and where they should target their attention. Likewise, on the hirer side, great to have the tools, and we're building them and to elicit what the hirer needs. The question is who are the people out there, how do you target them and how do you get the best person possible. That depends on marketplace data, not on conversational interfaces. You need both. But with the interfaces without the data, you just cannot deliver that experience. The last thing I'll say and that Kendra will talk about this more a bit later, because we've learned so much over a long period of time, we know which investment works. We do more of that less than the other stuff, and we retain our confidence that we continue the investment we need within the target cost envelope. With that, I'll hand over to Grant to carry on for us.
Grant Wright: Thanks, Ian. So Slide 11 speaks in a bit more detail to what Ian mentioned about how unique data and our AI capability is driving our 3 strategic priorities of placements yield and operating leverage. If I start with placements, our marketplace, as Ian said, is fundamentally about facilitating 2-sided trusted connections. The candidates trust us to deliver relevant results, but also provide realistic feedback about competition and their chances of getting the role. Hirers always want to understand what's out there, but also who's available and interested right now. That distinction between what looks good on paper to one side versus what's real and agreed to on both sides is critical to reducing wasted effort creating placements. And as AI noise increases in the market, this becomes even more important to generate placements. That's where we have unique data that cannot be scraped to DMZ. We see close to 1 billion and decisions on our marketplace every day. So that's candidates becoming active and being open to looking at opportunities. It's the jobs they review and save. It's what they apply to and then the jobs that they prioritize as their preferred applications that they're very interested in. It's the criteria that hires want to target on. It's who they search for, who they review, who they shortlist and who they contact. That data drives better matching, which leads to more activity, which leads to more data and therefore, drives better matching. So these actions that reveal people's real availability, intent and priorities drive those placements and create even more activity on our platform. That same 2-sided data also enables us to sustainably grow yield because we see real hiring activity and competition. We can predict, understand and manage performance and price. Our pricing system monitors supply and demand for about 45,000 labor market segments across APAC. That data enables us to provide attractive options for hirers to pay for more performance and the choices they make between those options and the performance they want gives us data on willingness to pay, which allows us to then further improve our pricing models. Lastly, on operating leverage, as Ian said, we've been investing in for over a decade. So we're not starting from scratch here. We've got advanced capability in building, deploying and optimizing AI models as well as a strong focus on responsible AI to make sure that we manage risk and retain trust with our users. And we're continuing to experiment and test to ensure that the innovations and algorithms we deliver have real customer impact. So that combination of existing capability, including applying AI internally for efficiency and then the experimentation to make sure that the innovation we're delivering is delivering real customer value enables us to deliver placement and yield growth while maintaining our operating targets -- operating leverage targets. Slide 12 then demonstrates how our proprietary data and AI capability are creating real value today, particularly through personalized matching and AI targeting. Our key inputs on the slide speak to the proprietary data that I talked about previously on both sides of the marketplace. The conversational AI products, live tracking of actions on our platform, verification through CPaaS and reputational products like reviews and reference checks give us a complete view of the requirements, preferences and importantly, the trade-offs and willingness to pay that both sides are willing to make a match. We then use that data in a wide range of AI models. So that includes proprietary ML models that we build in-house. It's LLMs fine-tuned on seat data and it's new technology emerging like sequence models, which is essentially the concept of a large language model that applied to behavioral data. So rather than predicting the next word, aims to predict the next action. We're also experimenting with agentic search systems to further advance our search capability. These prediction models are then used across the marketplace to deliver capabilities, including recommending targeting criteria to hirers based on what we know matters, to predict and manage performance outcomes and set our prices, to deliver highly personalized matches to candidates that manage both what they're interested in and what they're a good fit for. Our increasing ability to predict fit also allows us to target high candidates through exclusive outreach products for advanced and premium ads. And then we explain those matches using generative AI on both sides of the marketplace to increase confidence and encourage people to engage with the opportunities that will drive a placement. And it's this combination of the unique 2-sided data and our AI capability that's continuing to show up in increased value for customers. So candidates are now 1.5x more likely to see and apply for relevant jobs due to the quality of our matching and explanations, and that's underpinned our placement share growth of 7 percentage points over the last 3 years in ANZ. On the hirer side, hires are seeing the benefits of increased performance in advanced and premium ads by high targeting and they're opting to pay for increased performance, increasing depth adoption, which has grown 2.7x over 3 years and underpinning our ability to grow yield sustainably at 15% compound over the last 3 years. So we're seeing these results really show up for customers in our investment in AI. I turn to Simon on Page 13 to talk about how that shows up across our product suite.
Simon Lusted: Thanks, Grant. Moving to Page 13. I want to take this opportunity to drill into a little bit more detail on how this AI and data capability is showing up in examples of actual features delivered on SEEK over the last few years. We've broken this into the candidate side and the hirer side. What we're going to talk about is all live product in market that's been delivering for candidates. I want to give you a bit more flavor of the type of impact we're seeing from these capabilities. And our long-run investment in AI infrastructure, we were pretty really to integrating LLMs and natural language search into our core job search discovery experience. That delivered immediate uplift, which continue to compound as Grant and his team retrain models on our data today. But perhaps more important, it put us in a position to experiment pretty aggressively over the last few years with a range of different conversational experiences. We have 2 pretty exciting experiences in market right now, and we plan to use this learning to evolve and improve our candidate search and discovery experience over the next few quarters. While we were doing that, we also saw the opportunity in these capabilities just to do more of the work on behalf of candidates. So we launched our intelligent career feed, which uses candidates profile data, their behavior, all the data Grant talked about to do more of the reasoning on behalf of candidates. And that's given us a really big lift in the number of applications that come from this very low effort engagement experience. In fact, it's less than half the effort to find a relevant job through using our career feed than it is in manual job search. In addition, as our recommendations have become more precise, it's putting us in a position to talk to candidates off our platform with much higher fidelity, much higher cut through. We've really driven huge improvements in our -- the way we notify candidates, and that's allowed us to tap into monitoring job seekers who perhaps aren't active enough to be on SEEK at that time, but are willing and open to engage in new opportunities. That's delivered a 2x growth in channel performance in just a couple of years. And as Ian mentioned, it's not just whether the job is available, it's whether that job is right for you and whether you're a fit and whether the person on the other side is likely to want to shortlist you or progress you. And so we've been doing a lot to help explain to candidates how and in what ways they might be a fit. This has been particularly pleasing because what we found is that many candidates are a little bit hesitant to step outside their frame of reference. And when we can share with them that they actually are potentially a strong and high fit for this role, we're driving much greater levels of engagement and application, which is driving placements. And as mentioned, we've been investing in trust for a long time now. We've always thought that the labor market was sort of noisier than it had to be. And we've got a big team in [ APAC ] over 100 people now. We're in every market across APAC, and we're really scaling our ability to verify trust with new AI models that allow us to add authority, check identity and really deeply understand who is real and whether what they're saying on our marketplace is a genuine claim. And as Grant mentioned, the new products like Strong interest where candidates are able to nominate a few jobs that are their particular priorities. That's all underpinned rather by our trust infrastructure. On the hirer side, we're becoming more of an adviser through the job ad posting process. We're using our market data to explain the marketplace, help hirers build more quality ads, reducing effort, and that's driving up conversion of new hires to job posting. We've made big step-ups in our dynamic job ad pricing accuracy. And what that means is we're better able to understand whether a candidate's actions and the ad they're buying will lead to a placement. That's given us a lot of confidence to align prices to value. Grant mentioned our AI targeting. That's really been central to our ad ladder refresh. We've got an AI targeting feature that's bundled into the advanced ad and the premium ad, which really delivers a material difference in placement outcomes. And similarly, we're explaining to hirers why a candidate might be a great fit. So we're not only saving them time, but we're improving the chances that they make a placement by connecting them with candidates that they might not have considered otherwise. We're also getting further into the placement process. We launched in the last half our reference checking product, which is a full voice agent, a voice agent will interview a referee. It cuts the time to give a reference in half. It generates higher quality data, and it takes what was previously a messy offline event and brings it into our platform in a structured and actionable way. And this is the kind of data and the kind of trust that an intermediary can leverage to help both parties in a way that we don't think many can. So as individual features, I've tried to give you a flavor of how we're applying these capabilities to improve in many different places. But taken together as a system, I think I'm hoping to give you a flavor of how these benefits compound in each other. They lead to stronger, deeper understanding of our candidates and hire preferences, better matching not just more placements, but higher quality placements that drives ROI for hirers and that drives up their willingness to pay. So we're really excited about the progress we've made. But as Ian mentioned, we do still feel like we're just at the start, and we've got more opportunity opening up over the next few years, and it's an exciting time for us.
Peter Bithos: Thanks, Simon. I get the privilege of talking about how the data and products that Grant Simon just talked about, pull through into the actual numbers in each of the regions. And just a reminder, kind of the system that we're trying to build here is great data plus great products equal more jobs. And then you combine that with a great brand and great on-ground execution in every single country, and you hopefully drive results across APAC. And actually, that's what I get to talk about here today. So I'll start with ANZ. For those of you who remember last time, I kind of noted last time was actually first time where in a down market, we were able to drive ANZ growth and gain share. This time, actually, the results are even better. So it remains a down market, we were able to drive double-digit revenue growth, gain share, highest share in recent history in ANZ with a 17% yield uplift. So volume is slightly down due to macro, but share up yields and revenue growth well into the double digits. How that happened is explained on the bottom right-hand side of Slide 15, which I won't go into the details, but essentially, it was product driven. So a large chunk of the revenue and yield performance you can see is a dramatic shift in the types of products being offered and taken up by our customers. And those are the products that Grant and Simon just talked to. So you see a dramatic step-up in depth penetration, driving both placements and yield. If I go to Slide 16 and talk about the macros for a minute, it still remains a slightly down market. But I would say the biggest news positive on Slide 16 is the dash on the upper left, where we're seeing New Zealand turnaround after a period of substantial decline over 2 or 3 years. So we're pleased to see New Zealand volumes up on PCP. As for the rest of the slides, you can see the macro kind of stabilizing across ANZ from the previous few years. So good to see a relatively stable macro environment in an area where we can really drive share and drive the business pretty hard. That translates to now not just an increasing share, but importantly, an increasingly large gap in ANZ between us and the second largest competitor. So not only is it at record levels, but the delta between us and competition is now at record levels. And so whether it's the core variable pricing, the depth, the AI products, we're really pleased. It's a very strong result for ANZ. For those of you who have been following the Asia journey, would know we're going through quite a lot of commercial transformation, launching and rolling out freemium, which I'll talk to in a couple of slides. This half continued that journey. We had a full 6 months of Singapore, which we launched late in the second half, and we launched Hong Kong, our largest market. And I think if you kind of took a step back and you said if we were able to generate real revenue growth and launch 2 of our largest markets in freemium as well as have a full year effect of all the emerging markets and produce the results we have, we'd be very, very pleased. So we think it's a really strong performance. You can see paid volumes down. I'll talk to that in a little bit. Revenue up. Placement share as the survey calls it slightly down. But the survey, it's within the statistical variability and noise. We're kind of pleased with the long-term trajectory and all the other indicators that we crosscheck are really strong. And so we're pleased with that result. Paid ad yield, very similar story, and this is very consistent with what we've been doing since unification. Any benefits we're able to launch, we roll it out across APAC, and we seek to get those benefits across APAC. So you can see the advanced and the depth penetration equally shifting in Asia, just the same. You can see Slide 19 and then a few slides later on Slide 20 and 21, really tell the story of freemium as we go. So if you look at Slide 19, you can see the monthly ad volumes on the right-hand side. You can see the total ads increasing as we roll out the markets. You can see the mix shift slightly changed between paid and free between the pink line and the purple line. But then you can start to see now the monthly unique hires, which we've called out as a leading indicator that we need to get right longer term, now 18% up PCP, very pleased. So -- and then as we get more ads on the platform, the marketplace strengthens, we get more unique visitors, and you can see that in the bottom left. So strengthening of the flywheel in the Asian markets is occurring, and you can see it in multiple metrics on Slide 19. Slide 20 just talks to the placement and yield, which I talked to you previously. I'd note across APAC, for those of you that follow the placement over time, every 2 to 3 years, probably 3 to 4 years, we reset the sample size and also supplier. We'll be doing that. That's kind of a normal part of the cycle, and we'll reset the survey in the next set of results. So that's just a note for those of you that follow that. And then last, Slide 21 just talks to the progress in the rollout of freemium. Two weeks ago, we launched Malaysia. It was our last market. So we are now freemium in all markets. So when we get to FY '27, it's a clean full year. We're excited about that, but we're right now, I'm just happy with the results that we've gotten and shows the strength of the business and everything Grant and Sean spoke of. So I'll hand over to Kendra.
Kendra Banks: Thank you, Peter. Good morning, everyone. I'll begin with Slide 23 and talk to how all of this you just heard results in our finances for the first half. So we reported revenue of $601 million. That's up 12% compared to the same period last year or 11% when you exclude Sidekicker, which was not included in our prior period results. We delivered again on our commitment to operating leverage as total costs grew 8%, excluding Sidekicker, 3 percentage points lower than the revenue growth rate. This operating leverage is going straight to earnings growth. EBITDA was up 19% and adjusted profit of $104 million, up 35% for continuing operations. We reported, as you know, a $356 million impairment charge against our investment split across continuing and discontinued ops. The appendix slide outlines the details showing how the Zhaopin investment as accounted in our books went from $529 million in June down to $182 million this period. This is a noncash impairment and reflects changes in the last 6 months, which will help set the business up for a stronger future. Back to our core business. Our financial performance and trajectory are strong, and this provides the Board confidence to announce a record dividend amount of $0.27 per share. That's an increase of 13% from the prior year. Turning to Slide 24. Our commitment to operating leverage is clearly evident in these results. As I mentioned, excluding Sidekicker, revenue of growth of 11% exceeded total cost growth of 3%. We primarily focus that increased expenditure on our ongoing and growing investment in AI product and tech and the IT infrastructure and compute costs, which support the product experience. We fund this increased investment by being efficient with our run-the-business costs so that the total cost growth continues to be contained in our mid- to high single-digit total targets. As a reminder, we think about costs as a management team on total cost of OpEx and CapEx to create the right incentives. But of course, we do account for it split into OpEx and CapEx according to the accounting standards. The result for this half is a 7% increase in OpEx and 24% increase in CapEx, and that weighting isn't really a surprise with the prioritization towards grow the business activities, particularly the AI-focused product development that's delivering tangible results. Slide 25 talks to the drivers of adjusted profit, up 35% from last year. The growth was led by strong EBITDA performance, partially offset by a few below-the-line factors. The D&A was higher in line with the increase in CapEx over the platform unification years. Share-based payments expense increased. This included a one-off share grant to all employees and also includes the accounting standards requirement to value share payments on the date of grant, which happen to be close to a recent peak in our share price. Partly offsetting these were lower interest costs and a partial reversal of the Zhaopin performance fee following the impairment. Slide 26 is cash flow performance. I won't go through the detail here other than to say there's nothing unexpected. And as I said earlier, strong cash flows enabled higher dividend returns to shareholders. On Slide 27, our debt position was broadly stable. Our net leverage ratio continues to improve. It's down from 2.3x a year ago to 2 and is well within our target of less than 2.5x. And of course, that's driven by growth in EBITDA. Slide 28 outlines the performance of the SEEK Growth Fund. A reminder, we look at total portfolio value, including the portfolio valuation plus distributions from the fund. As of the 31st of December, this value was up 1% year-on-year. The fund's return on invested capital since inception is now about 33% with an IRR of 8%. There's more detailed appendix on the fund's 4 largest businesses as always. Also considering the fund, you'll have seen in the press today their announcement that they are commencing a process to sell their stake in Employment Hero. On Slide 29 is our capital management framework, which remains unchanged. Strong operating cash flows providing capacity to execute on our strategic priorities and provide dividend growth. Looking ahead, as you know, the fund opens a liquidity window in the 2026 calendar year, following which they must use reasonable endeavors to fulfill a liquidity request within the next 12 to 24 months. In the lead up to this, the Fund's Trustee Board is in active discussions on the optimal approach to maximize long-term value, and we will update you as the year progresses. Finally, briefly on Slide 30, our sustainability commitment continues to be a priority. Employment platforms like SEEK are uniquely positioned to drive fair hiring, and we're continuing to evolve our strategy and approach to expand our impact, strengthening platform controls using AI, working with other experts and organizations to continue to elevate fair hiring standards. Perfect. Ian to touch on the outlook.
Ian Narev: Thank you, Kendra. I'll be quick because we want to get to questions and I've learned over time, stick to what's on the page where we talk about guidance. The headlines are here, as you would expect in the half, we tightened the range. And as you will see on the right-hand side, that means for those of you who like to work to the midpoint, which we know many of you do. The guidance is a range, but it gives you a sense of where the midpoint ends up. And the only additional commentary on that, as you can see here is that the revenue and EBITDA is expected to be in the top half of the original guidance. So range contracted. We've done some on the midpoint for your benefit, but it remains a range, but we expect to be in the top half of the original guidance for the revenue and the EBITDA. So look, I think you heard here, I mean, I'll just quickly summarize a very strong result for us. There are no guarantees of success in this business, and we're going to need sustained hard work to maintain the momentum and to make the most of the opportunities. But we've got the foundations we need. And to come back to where I started, I think you'll hear a very strong sense of confidence from the management team based on the results we've seen and the assets we know we've got. And now it's up to us to continue executing. So with that, I'll hand over to Q&A.
Operator: [Operator Instructions] Your first question comes from the line of Eric Choi from Barrenjoey.
Eric Choi: So my 2 questions. The first one is just on monetizing placement share. So your placement share has obviously gone up to 36% now, but I think SEEK is probably only still doing 15% to 20% share of total placement costs in Australia. So my question is, can SEEK evolve its revenue model to be less reliant on total job ad volumes, but instead more on placement outcomes? Because I guess that might make you less impacted from any AI disruption of job volumes and it kind of still suggests there's plenty of scope to grow your yield over a long period of time.
Ian Narev: I'll quickly answer that, Eric, and then hand over to see if Simon wants to add anything. The short answer is if you -- this is quite important for the business. If you look top down at the amount any business is paying for successful placement, all our research shows we have plenty of room to move. Every bit of research shows that what we get for a placement is at the very low end of people's propensity to pay if it's a good match. That gives us a lot of confidence as long as we deliver the match. So we are nowhere near what we would consider to be a ceiling. At the moment, you can also see from here the combination of placement share and yield growth means there's a lot of money value we can create and value we can share through our current model. So we don't see an enormous need to evolve urgently, but we keep looking for ways to share the value with the customers. The team spend enormous amount of time on that. And at the moment, we're happy with what we've got, but we've got other things in the pipeline that we can experiment with and see how they go. I don't know, Simon, whether you want to add to that?
Simon Lusted: Yes. Yes, I do. I think we obviously think about this a lot, Eric. The key unlock for us for so much of our marketplace is knowing that a placement happens. I mean the real -- what we want to earn as much or generate as much value as we can when our customers make a placement and less when they don't. And so a lot comes down to our ability to predict whether a placement is going to occur given all the factors. And there's really 2 ways we're coming at that. We talked a little bit about that today. We've got great prediction engines. We've got the opportunity to play further down in the selection process. So we see more about how far candidates get. And that allows us to understand, well, how much value are we creating? And then the other way is just to play directly in the placement. We're doing that with guaranteed hire with SmartHire, which we talked about. Sidekicker is obviously in the placement. So we're thinking about these things. But the real unlock is the degree to which we can drive up our confidence in that placement probability. And we feel like we're making really good progress there, and there's a lot more to do, opportunities opening up for us in lots of different areas.
Eric Choi: Lusted, quick follow-up. I'm dumber than you. So can I just ask something a follow-up in dumb terms? Like everyone is worried about AI agents connecting employees and candidates directly. Just listening to your reasoning, like if there's a noisy world where employees are getting spammed thousands of applications by AI bots, it's -- you seem to be suggesting you're going to be giving higher quality matches probably because you've got, I don't know, intent availability data. So are you sort of saying your placements become more valuable in that world?
Simon Lusted: Yes. I mean I think what the matching is about sending strong signals of intent and quality. And if that gets easy to imitate at scale, then the role for an intermediary who can reinforce the rules, add trust and help candidates and hirers negotiate in a high-trust environment. That goes up, not down, granted.
Operator: Your next question comes from the line of [ David ] from [ Fabris ].
Unknown Analyst: Can I firstly ask about the ad ladder pricing model and strategy? I mean I appreciate you touched on it through the presentation. But can you help us understand whether you're running periodic price increases and then augmenting this with dynamic pricing? Or you're leaning much harder on dynamic pricing now and have moved away from those periodic price rises?
Peter Bithos: Yes. So actually, that's a really great question that gets to the heart of this result. This particular result, whether it is both Asia or ANZ, on about half of the yield uplift is being driven by product take-up of new products, right? So it's customers reacting to the probability to place expressed in, say, the advanced ad, saying, "I want that, I'm willing to pay for it," and they're making that decision by themselves. And our job is to educate the customer on the choices that they have. So this is not -- we continue always on the journey of variable pricing and making sure we price to the value we're creating. But this particular result is actually the strongest led by the products that we've actually created.
Ian Narev: I'll just add one point to that. Those of you who have been following since we did the dynamic pricing in 2019 know that the probability to place is a driver of the dynamic pricing model. So in the nonproduct-driven parts of it, the more confidence we've got in the propensity to place, the more the dynamic pricing model creates more value when we share in it. So even the nonproduct part, don't think of that as blood price increases. A big part of that is the data on the marketplace utilizing the dynamic pricing model to increase the value and the price because of the confidence in the placement. And that is a really important part of understanding the confluence of investments we've made over a long period of time.
Unknown Analyst: Yes. Got it. And then just continuing on, I guess, with the products here. Just that ANZ ad ladder penetration, I mean, if you look at the advanced tier, it looks like it's settling around mid-teens. And I think it was there in July, it was there in December and it was there through the period. Can you share with us where you think that optimal level of penetration may be and how you get there? Or on the flip side, what levers do you have to improve that penetration?
Peter Bithos: Yes. So that uplift -- so there's 2 things. One, we think there's further upside. Simon and Grant are working on new stuff, and we look forward to the customers taking that up, too, right? So like increasingly, depth is kind of part of the core revenue. It's not the thing on top. So -- and you can see that in the results. Advanced Ad in particular, pleasingly, we saw growth through the half, driven as our sales channels and marketing channels educated the customer and the product tweaks that we presented. So it's not a one-off. It's something we're driving into the business, and then we're looking forward to further improvements in to come. So this is kind of a -- we can do more of this, and we have very high confidence in that.
Ian Narev: Can I just add another thing for the benefit of a bit of historic context for questions you've asked for years? We've had questions for years, if volumes go down, do you get a procyclical effect on yield. Page 15 shows volumes down to yield up 17% and a really very strong performance from what we called the depth products. Now what does that tell us? Even when the volumes are coming down in this kind of economy, what we've learned is when these products transparently show a hirer that they've got a greater chance of a good placement more quickly, they pay for it. And we don't think that's anywhere near exhausted. And we're learning and there's more to go that it's probably less dependent on the economic cycles than we thought, and that's a really good message for everybody to a question we've all been talking about now beyond the 7 years I've been around.
Operator: Your next question comes from the line of Bob Chen from JPMorgan.
Bob Chen: Two questions from me. I mean just firstly, a clarifying one. Just looking at your cash flows, especially on the free cash flow line, it's obviously sort of come off a little bit. It looks like there is a bit of seasonality in there. Do we expect that to sort of normalize into the second half?
Kendra Banks: Yes. So cash conversion is always lower in the first half than the second because -- primarily because we pay out our employee bonus in the first half and revenue seasonally is slightly higher in the second. So that's what we expect to see. And the shift year-on-year is because we did pay a slightly higher bonus that came out of cash in this half than we had the previous year.
Bob Chen: Okay. Cool. And then maybe a more sort of general question around AI as well as that investment and monetization. Are we to sort of expect that the monetization of this investment in AI is largely through your depth of tiering as opposed to additional AI products on top?
Simon Lusted: I'll have a go at that. I think largely, we think, as Peter outlined, we launched a new ad ladder, and we said at the time, there's a lot of room for us to build new features, especially AI-driven features and enhance those ad ladders and drive more depth. So that is a big part of our future. In the last half, we launched a new add-on called assist. That add-on is really focused on monetizing the value we plan to deliver through things like automating reference checks, better ranking, helping people complete the hiring process more efficiently. I think that's more a long-term product lever, but we are trying to create a frame through which to monetize the efficiency elements of the AI work that we plan to do over the next little while.
Operator: Your next question comes from the line of Lucy Huang from UBS.
Lucy Huang: I've got 2 questions as well. So firstly, just looking into FY '27, how should we be thinking about the contribution from some of the growth drivers on yields, given like this year, we've had advanced ads, new product like next year, there won't be new advanced ads. So will we be leveraging more on dynamic pricing? Or do you think advanced ad penetration could still be a larger component of the growth there?
Peter Bithos: Yes. Lucy, it's Peter. I'll pull you up one level, which is instead of advanced ads specifically, the profit growth, which is we have new products and new constructs that allow buyers to get placements in various high confident ways. And we take those products and constructs and drive it into the base through our sales channels and our brand and our presentation through the product. That is a formula we want to continue. So advanced ad will be complemented by other things over time. But the underlying dynamic that you now have a system of for base price, we have sophistication in the way we price to value as we increase the probability to place. And we now have a system across 8 markets to produce products that are taken up in a very aggressive and fast way. And we're bullish that system will continue.
Lucy Huang: Understood. And then also recently, we've noticed you've got a people search tab on the SEEK Australia website. Just wondering if you can talk through how that feeds into the strategy for the company moving forward? And any early kind of statistics you can share on how many profiles there are, how often people are tapping into this database?
Simon Lusted: Great question. So we've got within the APAC Group, 45 million candidate profiles until now largely been only accessible to hires who purchased our premium talent search product. We've made the decision that a candidate profile should very much be at the core of the SEEK experience, and we can add trust to that. We want to make that profile more useful for candidates all through the hiring process. They should be able to share it with others, find each other, et cetera. So we made the decision to make candidate profiles free and publicly searchable for candidates who opt into it. We're really pleased with the rate at which candidates are opting in. It's really encouraging. And we think overall, it will not only strengthen our flywheel, candidates be more likely to keep their profiles up to date, invest in adding trust to it. It will drive and improve their job-seeking experience. But we also think that will drive freshness and depth, which will allow us to monetize through a more premium talent search offering, which we plan to launch in FY '27. So it's part of a broader strategic play to put the candidate much more at the center of our marketplace, not just the job, but the candidate, and it's going very well.
Operator: Your next question comes from the line of Entcho Raykovski from E&P.
Entcho Raykovski: So my first question is around the cost base and your comment that there are efficiencies in the cost base that are enabling greater growth investment. I wonder if you're able to quantify the extent of those efficiencies. I may be difficult, but is it sort of thinking 10%, 20% savings, sort of what that does to the velocity of product rollout. And I'm curious whether there's been a further step change over the past 6 months in those efficiencies with the developments we're seeing in AI.
Kendra Banks: Sure. Thanks, Entcho. So when we talk about efficiencies in the cost base, I'd point to a few things. The first is still seeing the benefits of the APAC unification, particularly in the commercial areas, where Peter's teams now run APAC sales and service marketing and all the kind of associated tools and corporate costs that support APAC, still seeing that benefit coming through in terms of tighter functions. The second is across the business in every function, we have very high take-up of AI as a core -- the AI tools internally as a core part of people's workflows. And we are seeing there -- it's a few points of efficiency in lots of different areas that's driving cost efficiency and allowing us to reinvest in the grow the business areas. And then in the grow the business areas, so product tech and AI, we certainly are seeing accelerated product development velocity. We haven't kind of disclosed a particular number there. But certainly, you can see it in the kind of pace of AI product rollout, while we are getting efficiency in terms of that product velocity flow and reinvesting it in continuing to develop our products.
Entcho Raykovski: Okay. And my second question is around the employment sell-down. Are you able to talk about the rationale for it given it's obviously a fairly weak underlying market for high-growth stocks in theory, might have been a better time 6 months ago. And then is the fund running an open process or do they have a buyer lined up already? I'm conscious that KKR was the buyer 12 months ago. So not sure if they're the ones who are looking to up their investment.
Ian Narev: I'd just say a couple of things. The decision to sell is entirely the funds. We don't control it. When asked our opinion, it's consistent with our goals as an investor to get liquidity over time, so we're fully supportive. Number two, the assets, terrific business done very well for the fund. It's obviously subject to value. And after this long and with this degree of understanding, they will sell if and only if they get a price that they think is good value. But number three, there's really no evidence at this stage that what we've seen in public markets, which has got all sorts of other drivers has translated into private markets. If it really has to this extent, that means the process wouldn't be successful. There's no evidence of that at the moment, and they'll find it out. In terms of who the likely buyers are, et cetera, that's something we have no visibility over.
Operator: Your next question comes from the line of Fraser McLeish from MST Marquee.
Fraser Mcleish: Great. Just a couple. Just firstly, on SME volumes. I don't think you've given your normal update on what percentage of your volumes are coming from the various customers? Have you got update on that? And then my sort of related question is to what extent, particularly in your SME volumes, do you think they are -- you've got unique listings that aren't going on to other platforms? That's my first one. And then just also, you've obviously outperformed -- you're going to outperform your high single-digit yield target again this year. You're just confirming that sort of over the cycle, high single-digit yield growth is still the target?
Peter Bithos: I'll let Kendra speak to the second question of the long-term yield growth. On the first question on SME, actually, when you get into the details, you're right, we didn't disclose it here, but partly because actually, there's nothing distinct or different in SME volumes to what the overall story in ANZ is. I'm assuming you're talking about ANZ. Actually, if anything, the depth penetration was a little bit stronger in SME. And the volumes are pretty consistent with the overall market. So share is up in SME, yield is up in SME, and we're very pleased with the results and SME wasn't differentially performing in any notable way.
Fraser Mcleish: Do we have unique ads?
Peter Bithos: In terms of our unique ads position, very similar to the flywheel, slightly strengthening, but again, not uniquely against our other segments.
Kendra Banks: Fraser, thanks on your question on high single-digit medium-term outlook for yield. We are still maintaining that as our medium-term outlook. We remain very confident, as already discussed, in the ongoing fast to future yield growth and that core dynamic that our customers are not that price sensitive when they're making a solid good placement, and we've talked a lot about how we plan to continue improving that delivery. However, as you know, we will always manage our pricing and yield to ensure marketplace health alongside. And therefore, despite delivering double-digit yield growth now multiple periods in a row, we're maintaining our high single-digit yield growth medium-term guidance there.
Fraser Mcleish: Great. Sorry, can I follow up very quickly just on that unique ads? I guess where I'm coming from is just with AI proliferation and having unique listings is potentially going to be more important. So I'm just wanting to try and understand the extent you have unique listings that aren't necessarily on other platforms.
Peter Bithos: Yes. So what we look at when we call unique listing, we then double-click and say, is it a unique listing that is directly on our platform versus pointing to a different place. So anybody can scrape and point to a different place and do that. Having said that, the large competitors in the marketplace don't scrape the direct ads from each other. And so we have a large corpus of unique ads that are available on SEEK, and that's remaining constant and consistent over the past few halves.
Operator: Your next question comes from the line of Tom Beadle from Jarden.
Thomas Beadle: Just my first question is just a follow up from [ choice ] on the cost side. I mean it's obviously nice to see that you've got your costs well under control and the positive [ choice ]. I mean if we take the breakdown of your cost base that you provided at your Investor Day, that growth in BAU bucket, I mean, what level of growth did you see in each bucket? And I guess I'm interested to hear any commentary you have around AI-related cost growth and just any other cost pressures in areas that are worth highlighting?
Kendra Banks: Sure. Thanks, Tom. So if we look at that run the business, grow the business split, I sound run the business for very low single digits, and then you can kind of do the math on the rest in terms of high single, low double in terms of our grow the business investment. And that includes all of the AI cost that we're facing into in terms of the cost of the team, the models we use, the compute cost, et cetera. So we still feel confident that the AI cost growth that probably is noticeable in the coming years, we can, at this point, confidently say is within our total cost growth target.
Ian Narev: And the other thing I think is your question, I'm looking at Grant here. We have under Grant's leadership, a protracted long-term piece of work just on using the same understanding of technology that drives our customer-facing platform to look for productivity opportunities inside SEEK. I mean you might want to just talk about a couple of areas, Grant, that we're really focused on.
Grant Wright: Yes. So Kendra mentioned supplying and encouraging AI tools for everyone at SEEK through their individual processes. We have an internal AI tool for that. We also use Glean, which provides enterprise search and agents and our staff are running about 5,000 agents a month at the moment to help with those individual tasks. The bigger opportunity we see is really reengineering the big processes in the business. And so we now have a team going into those core processes to help people map define them and do good process excellence work as well as bring those AI tools into that. So that's in train as well as then looking at our big opportunities for transformation in sales and service and product development and AI coding. What you tend to find in AI coding is that you can get big efficiencies today in writing the code, but writing the code is only a proportion of that cost base. So we're also looking at how we reinvent that process to become more efficient. Some of the examples of things we're doing sales meeting prep agents to drastically lower the time to pull the information and have a high-quality conversation. So raising the quality of conversation across the board and reducing time and effort. Real-time customer feedback monitoring and products. So we just get much more insight to all our product managers about what's happening over time, which previously was a big time sink and also meant that you couldn't get all that feedback synthesized. So this is showing up all over the business in these sort of processes. We've automated credit checks for our external things through agents. So there's lots of examples of more a process opportunity, and we really want to, in the next year, take that to our big processes and think about how we reengineer them with AI.
Ian Narev: The other thing I'd do just to add to that, our beloved Head of Engineering, James Ross, this is what I did in the holidays video that he sent to a bunch of us was a 1-hour video showing him experimenting with the latest developer productivity tools. They're very meaningful. So all these evolutions and revolutions on coding productivity, which the market is sort of looking at a gas and wondering whether they will disrupt our business. Well, we've made clear on the customer side that the data is the advantage. On the internal side, we've got people experimenting with these things as they come out. And as Grant said, we now want to map that to a really good process capability, but we think the opportunity is very meaningful.
Thomas Beadle: Great. And just, I guess, the second question just on volumes. I realize there's a little bit of caution in your commentary. I mean, can you just give an update on what you're seeing on volumes over the first 6 or so weeks of the second half across your markets?
Kendra Banks: Well, our January SEEK employment report will come out, I think, in a couple of days' time, and we'll give that insight. But as you can see from our guidance, at least on AU, it's very likely that volumes stay about where they are for the foreseeable future, and that's built into our guidance.
Operator: Your next question comes from the line of Nick Basile from CLSA.
Nicholas Basile: First question on placement share in ANZ. I think there was a stat there that SEEK now has a 4.9x lead versus the nearest competitor, which is, I think, up nicely on some more recent data points. So I'm just interested if you can sort of help us crystallize what you think the key to that result is and how we think about the sustainability of it going forward, I guess, particularly given you're telling us you're going to reset the data points or the survey data you collect, would just be helpful to understand how you're thinking about it.
Peter Bithos: Yes. So the key is actually everything that Grant and Simon talked about in their part of the presentation. So effectively, the product and the platforms and you can kind of look at it as a line item product like Advance where the platform overall as the system all of it is getting better. And it's getting better differentially against all other options in the market. And pleasingly, in our recent results, it's getting better against our next largest competitor and so all of that is happening. And it's happening because of the underlying data and AI and the capabilities. And because it's happening because of that, there is nothing that we would see would change that trajectory.
Ian Narev: I would just add with the normal point that we make, look at the placement share over a number of periods. And particularly now we're going to refine the methodology, we'll see what happens. It's a great story, both in Asia and ANZ. We probably are concerned less about movements half-on-half, more over 3, 4, 5 periods. That's what we'll continue to look at, but the signs are very good.
Nicholas Basile: Okay. Great. And a second question, just on the growth fund. I guess curious how you are sort of rating the performance of the growth fund. I think the IRR of 8% perhaps wouldn't necessarily suggest it's adding a huge amount of value relative to other investment alternatives. But of course, interested to know how you think about perhaps the intangibles associated with this strategy that investors need to consider, for example, playing in the start-up space is an important strategic consideration for SEEK and/or somewhat of a cost of doing business.
Ian Narev: Yes. Look, I think the IRR in and of itself looked at in the absence of anything else, is probably less than many people might think it would have been against the benchmarks of what other like funds have done over that period of time. It's actually not bad at all. And we have the capability inside, the team is very strong. So that is what it is. Interestingly, and I'll just remind people, in 2021, we were fielding calls about the valuations that the assets have gone to the fund, the fund is going to have a windfall and we said no. These are very fair market prices at the time. It was always going to be hard to earn the carry, and it's been hard to earn the carry. But we feel that the team is doing a very good job of it. And as a major investor, we're very confident in them. I don't think the informal connections are in and of themselves a rationale to be in the fund or not in the fund, but we talk a lot, and there's a very helpful and healthy 2-way dialogue on that, which I think is good for both us and for the fund.
Operator: Due to time constraints, we will now end our Q&A session. I will now turn it back to Ian. Please continue.
Ian Narev: Yes. Look, again, thank you all for your time. I think you've heard the main messages. The only thing I'd like to say just at the end is that there are a couple of people who have contributed to this result, who will be departing. Number one, Graham Goldsmith has done an absolutely outstanding job for all of our shareholders as Chair of SEEK and from a management team has provided a really great balance of challenging us and supporting us, and we will really miss him. Luckily, the Board has done a very good job in finding a successor. And likewise, as much as people know, it will pain me to say so, Dan McKenna has really done a great job as our Head of IR, and you've all interacted with him. This is his last result, and he's been a terrific contributor. So we want to thank him before he heads off overseas. And again, in Pat, we've got a ready successor standing in. So I just want to acknowledge the 2 of them. Thank you all again for your time, and we'll no doubt catch up with many of you over the coming days.
Operator: Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect.