Autotrader Group plc (AUTO) Earnings Call Transcript & Summary
November 5, 2026
Earnings Call Speaker Segments
Nathan Coe
executiveHello, everyone, and thanks for joining us for Auto Trader's results presentation for the half year that closed on 30 September 2026. Sitting with me today is our CFO, Jamie Warner, who will take you through the financial performance and provide a market update, after which I'll cover how Auto Trader is evolving beyond a marketplace into an active, specialised AI-powered platform that serves car buyers and retailers via Auto Trader Intelligence. When the presentation finishes, Jamie and I will answer questions from those in the room. The opening six months of financial year 2027 unfolded in line with our expectations, demonstrating the resilience of our marketplace, the significance of what we deliver for retailers, the strength of our brand, and the discipline shown by our teams. I'd like to express my appreciation to the entire Auto Trader team, as well as our customers, shareholders, and the broader group of stakeholders, for the trust and dedication they continue to show. Turning to Slide 2, which presents how we've performed over the long term on the key financial metrics spanning the past ten years. As the chart illustrates, despite navigating a range of macro cycles, disruptions from COVID, and the supply shortages seen recently, Auto Trader continues to produce a consistent record, and doing so requires very little additional capital; this ten-year history of compounding growth paired with strong margins provides us with an ideal foundation as we move into our next phase of growth, bringing together the strength of our core model with Auto Trader Intelligence. Now, the headlines for the first half of the year. Basic earnings per share increased 5% compared with a year earlier, reaching 18.1p, supported by our faster share buyback program. As expected, revenue excluding vehicle and accessory sales was broadly unchanged at GBP 301.8 million. That reflected the lower run rate we had as we left financial year 2026, caused by losing some retailers and some softening in spend following feedback on our Deal Builder product rollout in late 2025. Comparisons are also harder after the conversion of a stock offer in June 2025. During the half, retailer forecourt numbers rebounded, rising from 13,600 at the end of March to 13,900 at the end of September. Group operating profit was GBP 198.5 million, with costs kept well under control. Our operating profit margin came in at 63%, and it was 66% if you strip out pass-through sales of vehicles and accessories. When it comes to how U.K. car buyers engage with us, nobody else is close. People spent twelve times as many minutes on Auto Trader as they did on our closest rival — and that rival figure combines four distinct automotive marketplaces. Of all the minutes consumers spent across U.K. automotive marketplaces, more than 75% went to Auto Trader. Close to 80% of buyers come straight to us, which shows the trust consumers place in us, their confidence in our brand, and the depth of experience required to buy a car with confidence. As AI-driven search takes shape, we are, by a comfortable margin, the most visible automotive marketplace, according to ProFound, which independently tracks Generative Engine Optimization, or GEO. Traffic coming from AI assistants is still small at present, sitting under 1%, even as we keep improving how site search works and intend to adjust as consumer habits and expectations change; we have likewise kept raising the share of users who are logged in on Auto Trader. By the end of September, roughly 70% of our product page views came from logged-in users. Across the 6-month period, cash from operations reached GBP 215.3 million, and we returned GBP 330 million to shareholders via dividends and share buybacks. At September's close, our debt stood at GBP 335 million, with leverage of 0.8x EBITDA. We are proposing an interim dividend of 4.2p per share, an 11% increase on last year. Moving to Slide 4, where our strategy is laid out. What's shown here represents a genuine change in how Auto Trader thinks about and uses artificial intelligence; during past results presentations, much of what we said about AI centred on disintermediation dangers or on the ways we already put AI tools to work at scale. By now we have spent over ten years working with AI in LLM since they became prominent. With that experience behind us, our view is that the opportunities are greater than the risks. This technology amounts to a step change unlike any in a generation. Much like the shifts from print to digital, from PCs to mobile, from data centers to cloud, and from analytics to data science, we believe this presents an opportunity to make Auto Trader more valuable for car buyers, for retailers, for employees, and for shareholders alike. It allows us to broaden our addressable market, to strengthen our competitive moats, and to unlock possibilities that previously could not be obtained. It expands the range of problems we are able to solve and the pace at which we solve them, all aimed at reaccelerating top line growth. This is additive to our current business model and activities, not a substitute for them. We continue to view stock, Prominence, Deal Builder, and annual packaging events as the main growth drivers going forward. In fact, the strategy we'll lay out during the presentation should, if anything, make these even stronger. For many years we've steadily compounded both revenue and earnings, and we believe the shift toward AI technology will allow us to keep doing that well into the future. Auto Trader is already in the process of becoming an active, intelligent AI platform driven by Auto Trader Intelligence. That platform runs on our cloud infrastructure, our AI coding agent, 57 specialized AI models, 200 million truck consumer events each day, and 15 billion proprietary vehicle data points. We're able to accomplish things that general-purpose LLMs just aren't capable of, and we'll remove the hard work and decision fatigue from purchasing, selling, and managing vehicles. As the slide shows, this strategy centers on three core customer pillars, and I'll go into more detail on those later in the presentation. Briefly, then: for car buyers and the agents working alongside them, we'll leverage both our existing tools and AI to provide the deepest, most specialized AI-enabled car buying experience in the U.K. — delivered on Auto Trader. We'll make it available within the AI assistant, and we'll also offer it for AI agents, ensuring that regardless of where a consumer's search starts, Auto Trader remains the definitive source of truth wherever and in whatever manner they prefer; regarding retailers, we're transforming our retailer portal into a fully AI native operating system. Rather than just producing leads and a large number of data points, Auto Trader Intelligence will slice through everything to provide the exact actions a retailer should take, along with automated workflows for sourcing inventory, managing inventory, creating listings for executing dynamic pricing and triaging leads. Regarding vehicle manufacturers, we've gone beyond simple display ads and are now creating a comprehensive new-car deals marketplace experience, which enables them to move inventory more effectively and profitably. For close to 20 years, we've been making substantial investments in our technology. And while things will certainly change as a result, we don't see that as a threat to our current margin profile; but prior to delving deeper into the strategy, I'll pass things to Jamie, who'll walk through the financials along with a market update.
Jamie Warner
executiveThanks, Nathan, and good morning to all of you. I'll start by summarising how the group performed financially on a consolidated basis in the 6 months to 30 September 2026. Group revenue overall came to GBP 313.4 million. Of that total, revenue when vehicle and accessory sales are excluded was $301.8 million, versus $301.2 million a year earlier. Vehicle and accessory sales, for which there is a corresponding cost of goods sold, amounted to GBP 11.6 million, a decline of 30%, as we concentrate more on operating purely as an intermediary instead of carrying principal risk. For the full year, we now anticipate vehicle and accessory sales of GBP 25 million, below the GBP 40 million we originally guided, and consequently cost of goods sold for the full year will likewise be GBP 25 million. Group operating costs fell 2% to GBP 117.2 million, which speaks to our ongoing discipline on cost. Our share of joint venture profit, reflecting our stake in dealer auction, rose 10% to GBP 2.3 million. At GBP 198.5 million, group operating profit was 1% lower, while group operating profit margins held flat versus last year at 63%, and also flat at 66% once vehicle and accessory sales are excluded. Basic earnings per share increased 5% to 18.1p, and EPS growth outpaced operating profit because the weighted average number of shares in issue fell 9% following our accelerated share buyback programme. Cash generated from operations was broadly in line with the prior year at GBP 215.3 million. Now I'll turn to revenue by our main customer segments. As we set out at the full year results, since over 50% of all lease deals now complete through Auto Trader, we report our results under a single operating segment. Total trade revenue came in at GBP 267.2 million. Turning to trade, which remains our largest revenue source, retailer revenue came in at GBP 250.8 million, down 1% compared with a year earlier, and the following slide provides further detail. Revenue from Home Trader and other trade rose 7% and 9%, respectively. Consumer Services revenue also increased 9%, reaching GBP 22.9 million. Within that segment, private revenue grew 8% to GBP 14.2 million, driven by healthy volumes of private adverts, a substantial portion of which were listed solely on Auto Trader. Motoring services revenue rose 12% to GBP 8.7 million, aided by stronger finance revenues. Leasing, manufacturing and agency revenue totaled GBP 11.7 million. Within that segment, revenue from our standard manufacturing agency business—previously disclosed as a separate line—rose 18% to GBP 8.5 million, driven by ad campaigns from newly arrived brands seeking to build U.K. market share. Leasing commission plus ancillary revenue totaled GBP 3.2 million, compared with GBP 4.9 million in H1 of 2026, reflecting fewer deliveries amid lower van and pickup volumes, along with reduced average commissions following several funders' move to a fixed commission structure. Turning to Slide 8, we lay out average revenue per retailer alongside retailer forecourt dynamics. Average monthly revenue per retailer, or ARPA, climbed 1%, equivalent to GBP 43, reaching GBP 3,037 per month. Turning now to the three ARPA drivers individually: pricing contributed GBP 117 toward ARPA growth, which resulted from our successful execution of the annual pricing and product event on 1 April 2026, including a like-for-like price rise. Product accounted for GBP 15 of ARPA growth, arising from embedding Buying Signals within our core advertising packages, partly offset by a less favorable customer package mix. We have recently completed package changes, and those ought to help deliver an improved second-half performance. Stock lowered ARPA by GBP 89. Live vehicle listings were down 2% compared with a year earlier. That drop was largely driven by smaller independent retailers cutting back on the number of slots they take, mainly because trading conditions have gotten tougher and because of how they responded to the speed and scale of our Deal Builder rollout in November 2025. In the first half, the average number of retailer forecourts advertising on Auto Trader was 13,762, which is 2% lower than the 14,080 we saw in the same period last year; despite that, the recovery in forecourts over the period has been encouraging. As Nathan said, forecourt numbers went up from 13,600 at the end of March to 13,900 by the end of September. Slide 9 gives you more detail on livestock and retailer forecourts. Average live car stock on the platform in H1 was 454,000 cars, a 1% decline year-on-year. Even so, total live car listings hit 464,000 by the end of September 2026, up 4% year-on-year. We carried out a thorough analysis of retailers that churned during the tough stretch from November 2025 through January 2026. The figures reveal that among the customers who departed the platform, we re-signed two-thirds of those still trading, and we continue to concentrate on engaging with the remaining one-third. Turning to leasing, we delivered 3,447 units overall. Part of the decline in van volumes came from our decision to commit to fewer vehicle purchases, which you can see reflected in revenue from vehicle and accessory sales, while pickup volumes took a hit from the elimination of tax benefits. Now looking at costs on Slide 10. Overall operating costs fell 2% to GBP 117.2 million, though if you strip out cost of goods sold, they rose slightly to GBP 105.4 million. People costs within that were down 2% to GBP 51.5 million. Average FTEs came in at 1,203, as lower headcount and smaller share-based payment charges were partly offset by higher underlying salaries. We put 32% more into marketing, reaching GBP 12.9 million, though that jump mainly shows how unusually low H1 of last year was — this year's outlay is roughly what we spent two years ago. Depreciation, amortization, the digital services tax and other costs all held steady versus a year earlier. As noted, group operating profit overall fell 1% to GBP 198.5 million, while group operating profit margins, and the margins excluding vehicle accessory sales, were unchanged year-on-year. Cash generated from operations likewise stayed level at GBP 215.3 million. During May and again in July this year, we amended and extended our debt facilities, lifting total committed borrowing facilities to GBP 600 million through September 2030. As of September 30, 2026, our drawn gross bank debt came to GBP 335 million. After factoring in GBP 15 million of cash and cash equivalents, net bank debt was GBP 320 million, which puts our leverage ratio at 0.8x. In the first half of 2027, we returned GBP 330.2 million to shareholders. Of that total, GBP 269.6 million went to share buybacks—roughly 7% of issued share capital repurchased—and GBP 60.6 million was paid in dividends on the FY '26 final dividend. When it comes to allocating capital, our strategy remains the same. We invest in organic growth, keep our dividend payout ratio at roughly one-third of net income, and use any excess cash flow along with debt capacity for share buybacks. We still anticipate completing approximately GBP 500 million in share buybacks during financial year 2027. And finally, the Board is recommending an interim dividend of 4.2p per share, representing an 11% increase. Now let me move to our outlook statement. First-half trading has been in line with what we expected. As a result, we are maintaining our full year guidance for group operating profit to come in between GBP 395 million and GBP 415 million, with revenue growth accelerating in the second half. I'll now turn to a few slides covering the U.K. auto market. Transactions for both new and used cars continue to recover following COVID. Across the six months in this period, new car registrations rose a notable 12%, reaching 1.12 million vehicles, and manufacturers newly entering the market gave a lift in retail as well as fleet channels. During Q1 of our financial year 2027, used car transactions were up 1%, amounting to 1.99 million transactions, as supply among younger age cohorts is steadily coming back. The long-term fundamentals of the U.K. automotive sector stay solid. The U.K. car park keeps growing, and with strong new car registrations, we expect that to continue. Over the past six months, retail pricing—which is essential to how profitable retailers are—has risen at a steady pace, and how quickly vehicles sell has stayed broadly the same. Auto Trader's competitive advantage has never been larger. According to comScore data, consumers now spend twelve times as many minutes on Auto Trader as they do across the four brands that together make up our nearest marketplace competitor—compared with nine times a year earlier over that same period, measured against those same rivals. On top of that, Auto Trader still accounted for over 75% of all minutes spent on any U.K. automotive marketplace. During the first half, our internal tracking showed a bit of a slowdown in both the amount of time people spent and their monthly cross-platform visits, and this was particularly noticeable in June and July. On traffic share, we've gained ground against our rivals and remain stable relative to the sites of our largest customers — which suggests these patterns are driven by the market rather than by anything particular to Auto Trader. We also track the degree of influence we hold over sales, using information customers provide us directly, and that number was somewhat higher than a year ago — leaving us fairly confident about the significant role we play in U.K. vehicle buying and selling. With that, I'll return things to Nathan, who will go deeper into our strategy.
Nathan Coe
executiveAppreciate that, Jamie. During the past two years, like almost every company in tech, a lot of our AI commentary has centered on easing investor worries about disintermediation risk, or on calling attention to smaller feature rollouts; that narrative has obscured the opportunity this technological shift opens up. General AI models are only as capable as the data and context provided to them. A generic LLM trained on publicly available web text is unable to accurately price a specific four-year-old EV, confirm its provenance, or execute a complicated regulated finance transaction on a dealer's forecourt. However, when you combine cutting-edge AI with Auto Trader's 15 billion proprietary vehicle data points, 200 million consumer events monitored daily, 57 domain-specific models, and our extensive real-time integrations throughout the U.K. automotive ecosystem, you genuinely unlock a new capability that's nearly impossible to copy and that can support the finest specialized AI-powered assistant for U.K. car buyers and retailers; this is not just slideware. All the foundations required for this are clearly understood within our company. They are established, productionized, and already delivering results for us. Across Auto Trader's history, whenever major technological change has emerged, embracing it has consistently been our strongest defense. Every one of those transitions has strengthened both how we compete and the value shareholders receive. That identical philosophy underpins the fundamental change now underway. On Slide 19, we reach the central engine driving our strategy, Auto Trader Intelligence. Grasping precisely what it is matters if you want to see where we are headed. It is not one feature, nor a chatbot, nor some third-party AI wrapper attached alongside our current website. Auto Trader Intelligence is our scalable AI platform at the center — an intelligent orchestration layer that powers every interaction throughout the platform. Fundamentally, it brings together more than 57 purpose-built models—machine learning and LLM ones among them—plus a platform letting us replace or adjust however many foundation models we like, so there's no vendor lock-in and we can pick whichever model suits each particular task best. The 3 pillars are what set it apart, since a generic model or a start-up simply cannot replicate them. As noted earlier, we hold decades of proprietary data on U.K. vehicles spanning everything from the moment they leave the factory floor until they're scrapped—covering exact fitted specifications, prominence, valuation movements, and a great deal more. Our manufacturer build data and taxonomy are integrated deeply into the market, along with banks and lenders that give us real-time finance approvals, 350 automotive technology providers, and direct connections to thousands of retailers. A lot of these integrations work in both directions, since a large share of the automotive ecosystem relies on our APIs for operational processes and decision-making. And as I said earlier, we capture 200 million consumer events daily, which gives us live supply and demand and lets our models forecast pricing trends, buyer purchase intent and vehicle turn rates in real time. None of this is available to anyone but Auto Trader. The key here is this deep architecture, and that’s the reason Auto Trader Intelligence achieves accuracy, speed, and trust at a level public models just can’t match. That same edge is exactly what puts us in position to evolve from an advertising marketplace into the top specialist AI system for car buying and retailing in the U.K. To seize this opportunity, we have structured our strategic choices around three core customer pillars. Car buyers and their agents come first. Retailer customers are second, and vehicle manufacturers are the last. I’ll go through each of them now in a little more detail. Starting with car buyers and their agents, where our aim is to embed Auto Trader Intelligence straight into the consumer journey and change the way people buy cars. Consumer search behavior is changing. We’re shifting from rigid keyword-based drop-down menus toward conversational natural language queries and AI-assisted discovery. Narrowing down vehicles will always matter for making a good car decision. Still, the way you do that is undoubtedly going to change. Our capacity to handle natural language better than other models comes from our detailed taxonomy that covers every U.K. road vehicle. This detailed vehicle data framework—which today powers nearly 200 distinct search parameters, covering everything from precise capacity boosts to particular insurance bands, running expenses, and lifestyle characteristics—is crucial. On a conventional website or mobile platform of the kind we've run in the past, the reality is that you could only ever expose roughly 30 of those filters via dropdown menus before the display and the user experience turned completely cluttered and unusable for a typical buyer. AI search eliminates that interface constraint. A shopper is able to type or say a complicated multi-variable request in ordinary English—for instance: finally a dependable electric family SUV below GBP 20,000 with cheap insurance and a boot large enough for a double buggy. What you'll get is a set of results that doesn't involve guessing or approximating—it delivers fast, comprehensive, and accurate results that a generic engine or an ungrounded LLM just can't replicate. That being said, Auto Trader won't be the only place where consumer discovery happens. We strongly believe it will also occur within third-party AI assistance and personal consumer agents, and that brings us to our agentic readiness and model context protocol integration. Rather than allowing external LLMs to scrape static, unverified, unstructured web text, or attempting to make that task technically hard for them, we've deployed dedicated MCP servers and API endpoints. Through those protocols, outside AI agents—ChatGPT among them—can pull from our live database in real time, reaching structured data built on web standards. What matters most here is this: regardless of where a shopper's car-buying process begins, Auto Trader intelligence remains the source of truth underneath. We provide real-time inventory, vehicle provenance, verified pricing, finance quotes and part exchange quotes, and we then route that high-intent consumer who is ready to transact straight back to Auto Trader so the deal gets finished. Once a buyer has chosen a specific vehicle, trust, anxiety and decision fatigue are the biggest obstacles standing between them and a completed sale. For most households, a vehicle purchase ranks as the second biggest outlay they will ever make, and working through hundreds of thousands of listings to arrive at one vehicle can feel overwhelming and easy to get wrong. That is exactly where Auto Trader Intelligence will reshape the journey, turning those 15 billion vehicle data points along with all our consumer data into immediate, objective confidence for every buyer. We already do this through vehicle highlights, instantly surfacing rare features like single owner status or practical attributes such as a bigger boot, together with our market-leading price flags. Our capacity to deliver this helpful reassurance inside a usable interface such as search has not been limited by the arrival of AI. But confidence is just one piece of the story. We're additionally enabling shoppers to take action themselves and complete a larger share of their purchasing process on Auto Trader — a capability that is likewise offered, and will continue to be offered, to AI assistants in a uniform way across every listing on our platform. Deal Builder connects online browsing seamlessly with dealership completion.
Nathan Coe
executiveThrough it, buyers can immediately see the value of their card, obtain preapproved finance quotes, and secure the vehicle right there via a refundable GBP 99 reservation. Since Deal Builder turned into the core of our platform's offering, adoption has accelerated. Once we addressed the concerns certain customers voiced early in the calendar year, about 9,300 retailers are now active on Deal Builder — over twice the 4,000 we had at this time last year. And we remain on track to bring every customer aboard by the end of this financial year. Leads coming through Deal Builder are still closing at double the rate of traditional ones, and more than 50% of deals wrap up when dealerships aren't normally open. Auto Trader Intelligence removes friction and fear from the consumer experience, turning passive browsing into high-intent transactions — which strengthens our moat and simplifies things for both car buyers and retailers. Pillar two of our strategy is making our retailer portal an AI-native retailer operating system, turning it into the AI assistant for car retailing in the U.K. Car retailing remains quite split up, and many dealers still rely on separate software tools to handle sourcing, pricing, inventory and leads. With 18,500 people using our retailer portal each day and 6.9 million daily API calls coming through Auto Trader Connect, we already act as the operational backbone for U.K. dealers. Our plan is to bring more Auto Trader Intelligence capabilities directly into customer workflows, which should reduce system cost and complexity, save staff hours and help retailers get better returns on their efforts. The key point is that Auto Trader shifts on a retailer's P&L from a discretionary advertising line to a daily operational utility that sharply enhances their business, which gives us a strong base for monetizing new capabilities and generating long-term revenue growth. With Auto Trader Intelligence, we've already demonstrated this through products launched over the last 18 months. Previously we described these as falling under the Co-Driver umbrella, and that umbrella now carries far greater ambition within this strategy. Our computer vision models reorder ad images according to buyer click-through data, and generative AI automatically produces thorough vehicle descriptions and identifies key selling points; as a result, advert creation time has dropped from 28 minutes to only 5 minutes. Propensity models built with machine learning and trained on millions of historical sales score incoming leads in real time. Buying Signals indicate whether a buyer is local, highly qualified, and ready to purchase. Nearly 800,000 inquiries have been flagged as high intent, and those convert into vehicle sales at twice the traditional inquiry average, which generates substantial insight and efficiency for our customers. Our third pillar broadens our role with vehicle manufacturers. The U.K. The new car market is going through structural change, pushed by electrification, the Z mandate, and new brands entering. Manufacturers spend more than GBP 750 million a year on U.K. marketing, plus over GBP 5 billion on tactical pricing and incentives — a very material opportunity. Many of you likely know that for some time now we’ve pursued a strategy centered on new cars. And every time we’ve advanced, that progress has appeared in stock, in retailers, and most of all in audience. Every week, over 600,000 people interact with new cars on Auto Trader. We've built solid ties with manufacturers, and via Auto Trader Intelligence we plan to leverage that to offer a fast-growing market—one now up against stiffer competition than ever before—an advertising channel that's cost-effective and delivers high conversion. No one else is positioned like we are to let consumers compare new car deals. Our finance integrations run deep—with brands, with their captive finance houses, with retailers, and with leasing companies. Given that more than 90% of new cars are bought using finance, we'll enable shoppers to weigh cash discounts, PCP finance, and PCH leasing offers next to each other, backed by transparent deal scores. On top of that, we're able to tap live demand signals generated by our over 80 million monthly visits, which lets manufacturers launch focused conquest campaigns aimed at shoppers who are in the middle of comparing brands — and those campaigns produce sales volume you can measure. This means we can give OEMs a conquest sales channel that's transparent and highly attributable, while also growing the advertising budgets they commit to Auto Trader. And as our data models keep advancing, we'll be in a strong position to assist them in optimizing their GBP 5 billion-plus of spending on pricing and incentives, in much the same way we already do for retailers. A large part of our strategic shift has been made possible because we’ve sharply expanded our capacity to deliver software by embedding AI deeply within our engineering organization. We’ve talked for a long time about how strong our platform is. It is robust, scalable and supports high-velocity software delivery. During the past 12 months, we have built an Auto Trader-specific AI capability that runs on top of that platform. This is a coding agent that works with any model, meaning it isn’t locked to one provider—that lets us choose whichever model is most efficient and effective for each task. We’ve also embedded into the AI all the context, skills and tools it requires to function well within Auto Trader’s environment and, when needed, to act autonomously inside guardrails that guarantee speed, safety, security, accuracy and scalability. It runs on our cloud-based delivery platform, which provides our 400-person product and technology team with model-agnostic coding agents operating across 2 complementary models. One mode is engineering-directed: it plugs into developers' IDEs and CLIs to give real-time help with coding and safety checks. The other mode is autonomous execution, which lets background agents triggered by events — wired into Slack, Jira and GitHub — find and carry out tasks on their own. That setup is built specifically for our environment: it layers in shared Auto Trader engineering guidance, workflow tailored to each task, and safety controls, and it links AI clients to internal engineering tools, so engineers can keep using the workflows and tooling already in place today. Our coding agent has finished over 20,000 autonomous engineering jobs since it launched in March. Developer adoption is up more than 40% quarter-on-quarter, and we're handling roughly 10 billion tokens each month. The key point is that thanks to our centralized infrastructure and the way we tune our models, our net monthly spend on AI engineering has stayed flat even though workloads have gone up dramatically, which means software delivery speed is no longer tied to what it costs to run the platform. What we've discussed today is not some far-off plan involving a bunch of experimental AI prototypes. Auto Trader Intelligence is live, in production, and already producing real value at scale today. Embracing the technology early and working with it has always been central to Auto Trader's success. That approach means we've consistently been in a strong position to gain from technological change rather than treating it as something to manage as a risk, and we view the present moment in exactly the same way. Our infrastructure is modern and cloud-native, handling billions of real-time market signals.
Nathan Coe
executiveWe currently have over 55 specialized AI models operating reliably in production, and many of them are built on top of the highest-performing open LLMs. Our proprietary vehicle data exceeds 15 billion points, which no public LLM or newcomer to the market could reproduce. On Auto Trader, logged-in users account for 70% of all activity, we capture 200 million consumer events daily, and industry-wide integrations are invoked 6.9 million times each day. We also have embedded distribution reaching close to 14,000 customers, along with 400 highly skilled product engineers and data science experts who are fully devoted to improving how cars are bought and sold in the U.K. Our own coding agents have now expanded their capacity, letting us ship software faster, smarter, and more efficiently than ever before in our history — which means the scale advantage Auto Trader once held has grown much, much larger. For close to 5 decades, Auto Trader has guided the U.K. automotive industry through each major technological shift: print to digital, desktop to mobile, physical to cloud, and data warehouse to data science. Now we're making our biggest leap forward yet, transforming Auto Trader from a passive listings marketplace into an active AI native operating system. We think this widens the market we can serve, strengthens how hard it is for rivals to compete with us, and supports steady growth in revenue and earnings for a long time ahead. Our core business is solid, and our data network, technology, and team let us do more than just adjust to AI — we can fully embrace it, transforming the way cars get bought and sold across the U.K. So we're feeling positive about what's in front of us. Again and again, history has demonstrated that we can lean into shifts in technology and deliver benefits for car buyers, customers, and shareholders alike. Thanks so much for being with us today. Jamie and I will now take questions from those in the room.
Andrew Ross
analystAndrew from Barclays here. Two questions, please. The first one, I suppose, is to push you a little harder on the H2 revenue outlook, since you haven't been very specific beyond saying it's accelerating. So could you help us a bit more with the moving parts on revenue in H2, especially ARPA about stock from here, though the product piece could be somewhat better as some of the package changes flow through, and then price as well? And the second question is the unavoidable one about FY '28. I appreciate you're raising prices to dealers in January, and not all the decisions will have been made yet. Could you share some perspective on which products will be included in the tool set supporting that conversation? And given where we are today, how confident do you feel about pushing toward mid-single-digit ARPU growth in FY '28?
Jamie Warner
executiveSure — on the outlook, the first point I'd make is that the bottom-line revenue figure really hasn't moved; that's revenue excluding vehicle and accessory sales — and clearly, the adjustment I laid out to accessory sales carries no profit impact. So on the whole, what we guided at the full year, and where consensus stands, we're comfortable with. But you're absolutely right, Andrew. On the detail, I'll admit I'm not fully certain. A few items lean slightly positive or negative. Retailer forecourts are likely somewhat further along than anticipated, which is clearly a very good thing. My guess is the exit rate is roughly in line with what we'd have expected, though it will obviously be somewhat more weighted to the first half. That improved retailer figure is slightly dilutive to ARPA. And when you add in those higher forecourts, which are typically smaller ones joining, you get that sort of ARPA dilution. So the product lever, while it appears very large in the first half, will come down. I'd say the guidance remains completely solid. In the same way, on the stock lever — even though it appears quite negative in the first half, consider that in the back half of last year it came in at negative 85. Every bit of that negativity landed in Q4, meaning Q4 was really closer to minus 170. So the negative 89 we reported actually reflects solid progress and momentum. I think there's strong confidence that it will get better as we move from the first half into the second half. And once more, the 30 to 40 is fairly well supported, in my view. The product lever is probably the area that's a bit softer than we expected, and that comes down mainly to the package upgrades, though we knew the changes we'd made would only be ready by the end of September. Those changes are now in place, so we feel fairly optimistic about better customer penetration in the second half, with more motivation to upgrade and the product lever improving from that GBP 15. So if you assume a slightly better forecourt and maybe a slightly weaker product lever, retailer revenue still comes out at the same number; beyond that, consumer services is definitely doing better than expected, better than my own expectations. Manufacturing agency is likely a bit stronger too. Leasing is probably somewhat weaker. So again, taken as a whole, it all — nobody revised their numbers, and I think that's fine since it all comes together at roughly the same point.
Nathan Coe
executiveOn the FY '28 outlook, as you’re aware, we don’t provide guidance for the year ahead. But regarding the event, I think the key takeaway is that the entire Galberter situation is now firmly behind us. We’re quite comfortable with that. So in a way, we’ve learned our lessons from it. We’ve listened to retailers, and that has genuinely shaped the products we’re discussing for next year, some of which are already in testing and going live. There’s no reason to hold those back for some magical moment in April; we’ll get them live as soon as we can. Right now, that includes WhatsApp as a messaging platform and intelligent call tracking, which you can think of as instead of getting a missed call, you get a detailed lead like you would online — what vehicle they’re looking at, the buying signals, all those pieces of information that come with an online lead, which means you can actually govern, track and manage the lead the same way you would others. Today, in many dealers, they simply don’t know what to do with those. You get on average 3 calls. That’s not 3 leads — who do you call back, and in what order. So it’s actually — while it may not sound like the most glamorous of products — practically very, very, very useful to a retailer converting sales more effectively. Then we’ve got a couple there. We haven’t started on these ones, which we would also go with Buying Signals and appointments. So appointments is linked to Deal Builder. Obviously, some customers said, I don’t want a GBP 99 reservation fee. We kind of listened to that, adapted to enable people to request reservations. Some of them would like the ability for them to request an appointment and provide some times. So you’ve got that in there, and then a further enhancement to Buying Signals. So it’s a pretty strong lineup. Deal Builder — the Deal Builder episode is well and truly in the rear vision mirror; so I think we revert back to what we’ve always said, that we expect these events to be, in total combination of price and product, around 5% to 7%, and there’s no reason for us to think that would change moving forward. I knew you were going to ask that.
Jamie Warner
executiveI mean, I think, I think when you look at the trajectory, we've been pretty much flat in the first half. What worries people for this year is that consensus sits at low single digits. For next year, I think retail revenue is just under 6%, and total revenues are at 5.5%. That's a good trajectory to be on, and we're pretty comfortable with where all of those numbers stand. I wouldn't suggest that we're here saying all of those are moving up. But hopefully, we feel there's good confidence that this is underpinned.
Adam Berlin
analystThis is Adam Berlin with Goldman Sachs. I have three questions, if that's all right. You just walked us through a fairly detailed presentation on the new strategy and where AI fits into it. Could you attach any numbers to that? Previously, you were posting high single-digit organic growth of 8% in your core Auto Trader business. Does this bring you back to that level? Does it take you past it? How are we meant to think about it? Or is it really more about what you have to do to hold off the competition? Second, you mentioned minutes being down this half, possibly due to the World Cup, but could you say a bit about leads? The volume of leads you send to dealers, is that still holding up? Or is there an issue where people bypass auto trading and go direct? And can you give us any confidence that this isn't occurring? Then third is Deal Builder. You mentioned how many dealers now use it, but I think it would help more to know how many transactions, or what share of leads, Deal Builder is actually generating today? And is that meaningful yet, so we can understand how important the products...
Nathan Coe
executiveI'll take the first one. No, we aren't providing long-term guidance on what the strategy means for our numbers. As you know, I'd call us pragmatists. Right now our focus is, well, let's reach the kinds of numbers Jamie talked about and then take it from there. The reality is that the revenue we're — what's implied in all of what we're discussing is a deep push into dealer solutions and dealer software, a market where we really hold no share — no meaningful share. We've built dealer systems. Obviously, we know how to do that, but it's also to support the activity on Auto Trader. What we're discussing goes far, far wider than that. Implicit in that is the fact that we'll likely offer choice. We're not talking about extending a hard bundling strategy — we're still going to do that. That's sort of part of the strategy here. And what we're discussing sits on top of that. So in that sense, yes, we expect it to be incremental, but by its nature it will be something you sell to a retailer potentially one at a time at good yields, and these things tend to build over time. So I'd be surprised if — I'd like to see customer penetration next year. I'd be surprised if it were large enough to move into meaningful numbers. But we're doing this because we view it as an opportunity to — a pretty unique opportunity. We’re able to build things far faster, and we hold significant advantages when it comes to creating B2B software, since we also operate the marketplace. So this is a strategy that goes beyond the existing approach. I apologize that I’m unable to provide the precise figure. Sorry, I should mention quickly that this isn’t about warding off competitors. We just — the competitive landscape. In certain respects, we opened up our platform so anyone could address the dealer software space. And we’re still going to do that, but we’ll take on a considerably larger role ourselves. So this is very much a proactive step rather than something we feel compelled to do defensively. Sorry.
Jamie Warner
executiveWhat matters most — and here I'm looking specifically at the audience — is that the key metric we monitor near the bottom of the funnel is sales influence. We actually receive a great deal of sales data from customers. After someone has purchased a car, they tell us, and each search on Auto Trader carries a base case with it. That lets us match very precisely whether that person looked at that particular car, or another car that was also sitting at that forecourt, and then came in — and we're able to attach lead volumes to that, which rose marginally half-on-half. So we're quite confident that this traffic, as you put it, has some seasonality and perhaps a bit of market movement, yet it doesn't appear to be affecting consumers who are genuinely in market to buy cars, who move through the funnel via Auto Trader and end up buying. Then, with deal builder transactions, we've obviously altered the product so that not everyone holds a full reservation. Some people have requested reservation. As you'd expect, those come through in greater volume and convert at a somewhat lower rate. The way we've — in fact this is how Deal Builder has always been constructed, and the deal itself is really more of a packet of data. You build a deal, and it contains a buyer, a seller in the dealer, and an individual vehicle, and then buying signals get attached to it. There might be a lead or a message, though there might not be. That's essentially what a deal is, and it has grown considerably richer — first, obviously, with reservation, and then with Buying Signals. So I don't want to dodge the question: fullback reservation is probably around 10% of a dealer's sales influence now. We saw 5 to 10. So we have seen some improvement there. But in truth, the volume of deals, as defined by someone requesting a reservation or simply engaging with the deal widget — perhaps doing a part exchange, perhaps expressing finance intent, or just dropping in a message — is actually a fairly significant scale in terms of lead numbers. Still, I think where we're heading is that every lead, together with the call tracking changes Nathan mentioned, will become a deal, and every consumer interaction where we believe you're close to the bottom of the funnel and we see you as high intent will carry that packet of data, which we then work out. You can't pass it across — it wouldn't be passed to the customer without consumer consent — but there are things we capture around this being someone genuinely interested in this car, and we personalize the journey, present relevant journeys, and ultimately aim to connect them better with the retailer. That's the direction we're moving in.
William Packer
analystThis is William Packer from BNP Paribas. If I may, I have a few questions. First, in hindsight, I would argue auto traders probably faced an underappreciated headwind regarding the financial health of your dealer customer base — specifically around gross profit growth and those cost pressures. The slides today presented a relatively benign view of underlying market conditions. My sense is that we're really discussing the overall market rather than individual dealers. There are fewer publicly listed dealers now. From my conversations with dealers, it sounds like conditions remain quite difficult for them. Could you address how your customers are currently feeling about gross profit and cost headwinds? And then on top of that, you actually covered the second thing I asked about...
Nathan Coe
executiveWell, when it comes to retailers, there will always be some performing strongly and some struggling. On balance, I talk with plenty of retailers, and right now they're really not faring too badly. Sales are solid, gross margin is solid, and trade prices are dropping at present. Retail prices are staying firm. So that makes it a really, really favorable market to operate in. For independents, if used cars are your entire focus, that's certainly true. I believe you're right about the headwinds, and all of that has landed below the gross margin line. By and large, independent retailers appear to be getting along just fine, honestly. So, a couple of weeks back, we held our customer advisory group—which is made up mostly of 2025 retailers—and we discussed the market. They were fairly subdued, and with that group, that usually signals that business is actually very, very strong; it's just how they tend to be. Still, I believe those cost pressures remain. Last year they were especially severe, driven by the living wage, minimum wage, NII increases, and interest rates. Of course, I don't yet know what has been decided today, but many retailers carry a lot of exposure to stocking loans. For them, that's among the largest costs they face. Also, should interest rates climb, that particular expense becomes difficult—an almost unmanageable cost to offset. Retailers therefore remain extremely cost-conscious. Still, the pressure surrounding it seems far less severe than it did a year ago. It bears mentioning that for franchise retailers, the key distinction is their entanglement with the new car market. How you feel will hinge on which brand portfolio you hold. If you carry plenty of the new entrants—whether you committed to them early or signed on over the past few years—those vehicles offer excellent value, sell very quickly, and deliver solid margins. Working with the manufacturers is genuinely pleasant. They respond fast, and in some ways they need less hand-holding than certain other partners. If you're operating heavily in the legacy space, a number of those older manufacturers are struggling right now. Competition is part of what they're up against; regulation is another part. Whenever pressure builds on the system, that pressure in real terms ends up being split between the manufacturers and the dealers. So among some franchise retailers, you see a bit of tension in particular groups that haven't properly arranged their portfolios. Still, few of them are publicly listed these days, and those that are seem to be doing reasonably fine. And I don't think that isn't representative.
William Packer
analystThat's encouraging. Just one quick follow-up. Regarding the AI products you offer customers on the B2B side, it sounds like you're pushing into that somewhat harder than before. Over a medium-term view, could your customers move to Auto Trader as their DMS and thereby unlock meaningful savings? Or is that a step too far?
Nathan Coe
executiveI think the term DMS, and we've separated this out ourselves from a strategy standpoint, is honestly a fairly awkward label for a whole range of different functions. We all do the same thing, by the way, so that's not a dig at you for using it. And that covers everything from managing leads to invoicing to booking after-sales. I think we feel we won't try to solve problems that are already solved well today. We're really focusing on the areas where dealers aren't served today, though some of that functionality is, in certain cases, supplied by those BMSs. So will we go into, you know, the Pinewoods and the Keyloops of the world, those really hard-core systems sitting right at the heart? No, because we don't believe that by doing those differently you can generate much more profitability for retailers, and that's really where our focus is; but some of those systems handle various levels of pricing, inventory management, online advertising, dealing with leads, not just Auto Trader leads, finance systems, and we've got a unified finance platform. So there's plenty for us to go after. And I think it's fair to say we've been laying the foundations, but we haven't necessarily leaned in as well as we're able to now, because we can build a lot more software a lot faster; and our platform means all the services are there. And building services, AI isn't as good at, but building things on top of services and interfaces and that sort of thing is excellent.
Analyst
analystFirst off, I want to ask about Auto Trader Intelligence. At what point should we expect it to be driving monetization for an agentic service or an outcome-based offering? Then a second thing: that 1/3 of dealers who churn and, as you mentioned in the prepared remarks during the presentation, are still holding out, will the package of changes you've pushed through from the dealer advisory councils be sufficient to win them back? Hopefully that makes sense. And with the stock available for reservation now getting to the stage where it feels like you could begin shifting the marketing message to consumers toward being able to make a reservation, are you thinking along those same lines as well?
Jamie Warner
executiveThere I run a bit of a risk of repeating something I said earlier. I think we clearly see more opportunity in what we've set out being stand-alone and being monetized at the point where we're going to begin, and we appreciate that whenever we develop one of these products or features, we'll still go through that same process of asking whether it's one of those things that benefits everyone or whether it's something that could be monetized on a stand-alone basis. So, as Nathan said in the presentation, we feel it can be incrementally pretty meaningful. I think today we're simply setting out where that opportunity sits. We're not here talking about FY '28, '29, '30. That will come down the line in future presentations.
Nathan Coe
executiveRegarding the second question, which I was jotting down while you were asking it, just to confirm I have it correctly, are we doing enough to win back the remaining retailers that basically...
Analyst
analystYou set up the customer advisory council, you announced that WhatsApp messaging was one of the — is that sufficient for those churned customers to come back?
Nathan Coe
executiveYeah, we believe so. There's a bit of art and science involved in figuring out who has churned and who has returned. Being a car dealer counts as a regulated activity; so understanding who has left, who has churned, who has exited the industry altogether. I mean, that was the majority of retailers when we examined the numbers — they've actually gone out of business; but I don't believe there's anything further we can do, honestly, and it feels like sentiment is definitely turning in our favour, while customer adviser groups number around 20 to 25, and the people in them represent 1,000 independent retailers. And what they've told us in response is that you've listened, you're changing things. And then it boils down to offering retailers the best sales channel; and I think if anything, we've seen just as many retailers opt in and return of their own accord as opposed to us winning them back, because they've tried other platforms and discovered they get more sales faster on Auto Trader. And in that regard, from a competitive standpoint, which Jamie highlighted, it's stronger than it's ever been; so whether the retailer count goes back to dipping a little because conditions get tougher and the industry returns to its longer-term average, who knows. But I think we ought to be able to return to that same number of retailers with the same level of penetration that we had back then. I don't think there's anything else we can do. I think we genuinely — this and we can — we change things. We're launching products. We've engaged more widely. We've always engaged with retailers heavily. It was the smaller retailers, who are quite hard to reach, that we're now doing a better job with, I think. There are — I should mention actually, there are a few areas. So if you're a small retailer selling very cheap cars, it's worth recognising that the Auto Trader rate cut doesn't really work on value of cars. So for some of those retailers, actually, the proposition might not be as economic for them, because the effective cost per sale as a percentage of gross margin is higher. So we are examining, in parallel with many of the changes we've discussed, we have an economy stock package available to reach some of those retailers — still lots of retailers will advertise those cars. But if you're not going to use Auto Trader, that might be one of the reasons, and there's a bit of an economic reason behind it. So we're looking at — there's a lot more stuff we're doing in parallel that I think will help us get there.
Jamie Warner
executiveOn the marketing side, then — yes, I believe there's additional room for us to focus on promoting more from the consumer angle, the journeys we offer on Auto Trader, though nearly all of that is bound up with what we're pursuing through Auto Trader Intelligence to carve their agents, adding personalization, listings that are better and more relevant, smarter search journeys — and every bit of it concludes with that deal packet I mentioned, and there'll be the choice to reserve, request a reservation, grasp financing options, part exchange, possibly some customers' appointments, as Nathan mentioned. So we're definitely beginning to evolve the consumer experience so it speaks more to those bottom-of-funnel experiences. And in the end, what we're aiming to do with it is bring more highly qualified buyers to our retailer customers. Everything concludes with the dealer, everything concludes with their ancillary products and how that complements what they're attempting to achieve when it comes to gross profit.
Lara Simpson
analystThis is Lara Simpson at JPMorgan. Apologies, but I have one more question, this time about the DRAM retailer figures. Clearly you've discussed those that came back, but I believe among the ones you last — you mentioned roughly 54% are no longer trading as well. So could you expand a bit on that closure story, or consolidation, and on whether you sense retailer consolidation is turning into somewhat more of a risk as we go forward, and how that plays out for retailers, not in the second half but rather into '28 and beyond, both regarding absolute retailer numbers and how that consolidation might contribute to a weaker product or price mix? Moving to my next question, which is about margin. It's clear you've held the core margin steady. My sense is that people costs have come down somewhat, and that's helping pay for part of the marketing spend. You're plainly telling us the intelligence platform pays for itself. From a midterm view, how should we think about the pieces that make up margin? Is the goal to keep it where it is, and maybe reduce headcount and people costs a little more, so that funds the bigger step-up in marketing and AI? I'd just like to understand that. And my last question, sorry, is just about Amazon Auto, which I take it adds some uncertainty as we go into the new year. What I'd like to know is how you see the launch timing playing out, and whether the chance of them coming into the market over the medium term is shaping your approach to pricing talks or next year's marketing plan — just your thoughts on that.
Nathan Coe
executiveI'll handle the first question. Alright. Regarding the dealership, to put it briefly: those closures don't look unusual to me, and I don't regard dealer consolidation as an increasing risk whatsoever. That's never been how I've seen it, and honestly it hasn't really arisen in that light. Here's what we're really seeing — plus the underlying cause: entry barriers are simply extraordinarily low. Earlier, there was a chance for some consolidation among franchise retailers, driven mostly by OEMs shrinking their footprint, but that opportunity has now been fully taken up. We've got no room left for franchise dealers, since a lot of the new Chinese entrants want to move product in a very traditional manner. I'm not using "traditional" negatively there — I believe that approach is precisely the correct one. Even across recent years, when you consider the largest dealer groups, my sense is that a few of them will end up larger. Yet whenever we've watched certain companies expand, others have divided and carved off major portions—something that ties back in part to the pressure I mentioned in my conversation with Charles. You may find—actually, it could have been < >. Should you own those older manufacturers, it may be worth revisiting the way you think about your portfolio, and a number of groups have openly removed particular brands from their offerings. So if anything, what you gain in one place you lose in another. My recollection is that dealer numbers tended to fall about 0.5% annually. That was prior to COVID. We would talk about it. Nothing we observed struck us as structural or hinted that fragmentation was on the rise. These closures don't get much discussion from us, but whenever you look at — honestly, whenever you look at the Auto retail figure, it appears remarkably steady at the top; yet when it comes to independent retailers, and sometimes we can't even tell which ones are exiting and returning as the same business. Openings and closings happen constantly. And I don't believe during that stretch it was far out of line with history, even though trading has become — COVID was a fantastic time, truly fantastic. Everyone was earning money; it has gone back to normal. So it has become somewhat harder, and you were hit with a set of costs related to NII and minimum wage. So — and stocking loans. So perhaps during that stretch it was a bit more severe, but at present, I think churn levels are back to — everything has returned to normal.
Jamie Warner
executiveThe attrition figure for the opening six months of this year came in at nearly the identical level as the opening six months of last year. Almost right down to the dealer. At the margin, so the confidence regarding the long-term margins. < >, we have already discussed during this call that revenue growth is improving. So that is a help. So there is greater capacity to absorb some cost growth. Headcount went down during the first half, though it is really worth pointing out to people that the bulk of that reduction stems from the leasing integration. So it has nothing to do with engineers, developers, data scientists, anything of that kind. Those are fairly consistent; and I think I would model across the coming 2 to 3, 4 years, it is always quite difficult to know precisely, but I think headcount fairly flat and some salary increases as average salaries edge up. Marketing should be dialed back a step; that was really only because last year was very low. It will be in line with revenue growth; and then we have always for pretty much close to the best part of 10 years, those other costs, we do have a genuine process around how we manage them, and we really make sure that we are extremely disciplined about what is going on there and how we manage it; and I think the propensity for us to do that, we feel it is one of the areas where we feel confident. And hopefully, we demonstrated in this first half, there is virtually no revenue growth and margins have not moved. So yes, I think we feel okay about it those are the moving parts.
Nathan Coe
executiveSo with Amazon Autos, we're not sitting back at all. We keep an eye on every bit of that. I'd say we have a strong business, and that's part of the reason we've done as well as we have across the years, though I'd like to think some of it also comes down to sound judgment and watching these developments closely. What we've — what we understand about Amazon is that they excel at new goods, at consumer marketing, at cross-selling, and at items that fit in envelopes and that can take advantage of their logistics warehouse. A car transaction, though, is a really big metal thing where finance gets sold, and it's used. So each one is unique, and usually a part exchange is involved too. So any time we examine these competitors, we focus almost entirely on the consumer side and on what could solve a problem we aren't solving today. Then we sort of get on with making sure we solve that problem if we weren't already; in this situation, we don't see what that is. There's a trust element they're working on — I mean, they'll be here and they'll launch early next year; we're fairly confident that's highly likely. But what they can offer consumers is a bit harder to discern. Launching one of these propositions with retailers is very easy. You simply tell them it's free and they'll hand you some cars, particularly if you're a brand like Amazon. Yet the only thing that ultimately matters to a dealer is car buyers. They've been in the U.S. now for, I believe, 2 to 3 years, and we're pretty well connected with people over there, and it isn't something people really discuss as doing a lot; it might actually be that they're pursuing their advertising opportunity, trying to gather more data on people's vehicle interest. And honestly, I could see that potentially being a good lucrative space for them, but that's more a Google, Meta, Amazon space rather than anything we're playing in.
Jessica Pok
analystJessica Pok here from Peel Hunt, and I've got a couple of quick questions, if I may. First, I'd like to get clear on where things stand today: you have the ChatGPT app, but have you actually made your stock available to AI agents as well at this point? By that I mean — the U.K. was, I suppose, instincts in the U.K. as yet, but would you open it up to news if it arrived in the U.K.? My second question is just on — I think I may have missed this. Nathan, you spoke about what you're able to offer agentically at present and sort of down the road. So what's available today versus what you're planning for later on? My first question is about listings, vehicle check finance, and Deal Builder — which links in that chain can LLM agents access at this point? And the last one: clearly you're not going to say exactly when revenues can speed up into double digits, but when it comes to product launches, should we expect the rollout pace to pick up in the near term?
Nathan Coe
executiveI'll take the first question. Is our inventory available? Yes — through MCP connectors and app integrations within several of those AI assistants. Have we opened up our — at present, bots like Google are unable to crawl our inventory. And when you use real-time inventory inside an AI system, what those systems typically do is query a search engine. That explains why historically it hasn't. It doesn't deliver strong results for real-time inventory; it isn't fast enough. And quite often, those search engines lack access to listings; they definitely lack access to Auto Trader listings. What I think we're signalling is our stance on search engines and, more broadly, would we open up — if someone hits our search API openly without a password, would we simply dump the inventory? No, we'll be much more selective than that, because we believe the companies doing this well number 3, and those are the ones we should partner with, since they're the ones that will get used. We see no reason to take the risk of simply opening up Auto Trader to be used in a headless way, because with those custom integrations — which leads into your second question — we can pick precisely what we expose and the manner in which we expose it, mostly in fact to deliver the best experience for a car buyer. We've answered this before — Jamie raised it earlier when discussing the audience. Rest assured, our attention will remain fixed on safeguarding the lower part of the funnel. And we believe we'll likely do a great deal to guarantee that outcome, though my suspicion is we might not have to do quite as much as people anticipate, since once those interfaces are in use — at present Instinct is pulling data from Auto Trader, and frequently a cookie notification appears, which is definitely not the best user experience; yet even within something like ChatGPT, where you have custom integration via MCP, the results returned are substantially, substantially superior once you have the app in store that's now aligned at the moment you run the search, and those results are truly genuinely excellent. The logical follow-up, unless someone is prepared to purchase a vehicle based solely on a picture, is to select one and go deeper. Regarding the audience questions, we won't claim for a second that fewer people will take that route. Every one of us relies on AI assistance. Suggesting that won't occur would be absurd; however, the top of funnel has never been somewhere we concentrate. Many others could be affected by that; our real concentration is on the moment when you're prepared to decide. You want to view what is actually on the market for purchase. That is the point at which we get involved. That connects to your second question, and it mirrors precisely what we encountered with Google, the App Store, and earlier cases. The honest answer is we can't say exactly what we'll make available — we'll figure that out as we go. Usually, these choices can be reversed; in effect, you're able to open and shut features. My expectation is that we'll take a step-by-step approach to exposing more, learning how users respond and whether it delivers value for them. Really, this comes down in part to weighing the upside against the downside. A point that may have slipped past some observers is what took place once Google entered the picture: our audience grew larger — far, far larger — and our market share climbed far, far above anything we'd previously achieved. And in my view, AI assistants that help less confident individuals jump into Auto Trader at the proper point represent a substantial opportunity for us to connect with people who may not be using Auto Trader at present. And strange as it may sound, a portion of those people head straight to franchise retailers. For the most part, other platforms such as Auto Trader aren't part of their habits, yet a certain segment of users exists there—and by building on the efforts already underway, there's a chance to boost Auto Trader's audience strength.
Jamie Warner
executiveYeah, and regarding that third point you mentioned, there aren't any ground promises or announcements about revenues, though we feel pretty comfortable with where the consensus numbers sit for fiscal '27, fiscal '28, and fiscal '29. As for the second part of your question, 100%, the pace means you'll be seeing more products from us, and that really comes down to what Nathan was referring to with the AI coding agent, our capacity to ship more software has gone up pretty dramatically.
Edward Young
analystEd Young, Morgan Stanley. I have just one question, please; it's really a follow-up on the minutes topic brought up during the presentation and in the earlier question. Your visits and minutes have fallen, yet the gap to #2 has widened; so on the whole, clearly, engagement in car marketplaces is lower, and there may be some noise. More broadly, do you expect Auto Trader Intelligence to cut the time people spend on the site, perhaps because they arrive at the right moment? I'm only trying to figure out how minutes and visits ought to be tracked or interpreted as a measure of motor or a measure of overall strength — you brought up sales influence, and clearly that isn't visible to us. So just a few thoughts on how we should interpret the present market situation and those trends as time goes on.
Nathan Coe
executiveYes, both questions are good, but the second one is especially good. I think, yes, I mean, no one wants to talk about the weather and sporting fixtures while presenting results, but for whatever reason it was simply a bit softer across the summer; and that was a market thing. We're in constant contact with retailers. That showed up in later sales too. Sales have been really uneven over this stretch, and it's pretty common for a retailer to reach out to me two weeks in saying the month looks awful and asking what's happening. Then two weeks after that, things turn out completely fine — so it's been somewhat odd. You're right that we keep our attention on the structure. There hasn't been any genuine change; if anything, we've actually picked up share. So it really seems like a broader market issue, possibly connected to demand weakening. Some demand indicators are starting to ease a bit. Normally, that only shifts prices on used vehicles, and we’re already observing that in trade. Still, I want to be very explicit: we’re aware that people are interacting with the Internet in a different way—all of us are—and that is going to alter certain patterns. We simply don’t believe—in fact, we think it could end up making Auto Trader’s position stronger overall, not weaker. Determining what using AI on Auto Trader means for a particular metric such as time on site is an extremely, extremely difficult question to answer. And I think we will have nothing to hide on this. We’re going to build it out, and we believe it’s the right thing to do. Our emphasis is on people first and relationships second—those are the metrics we watch most closely. Think back roughly five to ten years, when lots of us discussed page impressions: the surest route to raising them was to blur images. Consider that approach—you can readily gain something near ten percent while time on site stays the same, merely by making servers respond more slowly or adding a couple of steps to a workflow. That is exactly why we have never optimized for it. So you might wonder why we mention it at all. What it comes down to is that engagement is what counts when we're up against our competitors, and when you look at a source like comScore, it's the strongest engagement measure available; so admittedly it has limits, but nothing better exists. And that, I think, explains why we pay somewhat more attention to the ratio and discuss how it appears at the bottom of the funnel.
Joe Barnet-Lamb
analystJoe from UBS here. Just 3 — no, I'm joking, just one question. So regarding your GBP 500 million buyback for '27, does that bring you to the leverage level you ultimately view as optimal? Or is it just the buyback size you considered rational for this year? Obviously, I'm trying to work out whether you're prepared to take on more leverage next year to buy beyond free cash flow.
Jamie Warner
executiveYes. I believe we mentioned it in the May results. Based on what we knew at that point and what we know today, GBP 500 million looks like the correct figure — that gets us to 1 turn of leverage. That isn't a ceiling or a target. If we still judged it the right call, we would go above that level. Obviously, a lot depends on many factors that come into play. But we're comfortable with the GBP 300 million we've bought so far. And as I said, we expect it to be the GBP 500 million.
Joe Barnet-Lamb
analyst< >
Jamie Warner
executiveThe last one at the back.
Analyst
analystApologies for joining a little late from the back. I have a few questions, if that's okay. First, regarding the OEM opportunity you described, Nathan, what short-term actions do you plan to take and which levers do you intend to pull to push that forward? And second, as the way people interact with Auto Trader changes somewhat, are you seeing any weakening in your capacity to attribute sales, or any shifts in how you need to collect data and engage with consumers and retailers to keep things robust...
Nathan Coe
executiveYes. A good question to finish on. So concerning the manufacturer opportunity. Retailer — it has 3 parts. We have leasing, plus our work with retailers and with manufacturers, and across those three areas we maintain distinct priorities. On Auto Trader, retailer and genuinely new car inventory is doing nicely and has done so for quite a while. It's basically about extending penetration among over 50% of franchise retailers, and we've brought in additional franchise retailers as well as fresh stock. That piece therefore feels relatively straightforward. I realize somebody here would tell me, Nathan, that's not exactly fair. Still, I believe we have a solid route forward in that respect. What's missing from those new cars — and I'll concede this requires somewhat more effort — is financing options presented next to the cash price for new cars. Hardly anyone pays cash for a new car. What's interesting is that a lot of people don't pay cash for used cars either, yet everybody relies on the cash price to figure out which vehicle they'll end up with. So for the 600,000 people who visit Auto Trade each week to browse new cars, that's still extremely useful, but we believe we can take it further by showing you what the real deal looks like, since quite often the actual deal on a new car is hidden inside the finance package, whether through a deposit contribution or APR subrogation. So there is — that's really — that emphasis on finance does in fact extend across every one of the new cars. Regarding the deals hub, we have leasing deals, which are basically a financing option. Our next step is making sure financing offers are attached to new cars that are physically in stock, so we can then compare the two across HP, PCP and leasing. After that, we can tell users that this is the one place where they'll find the best deal across all three of those measures. That's really about finance integration, and probably about bringing more retailers on board. As you can imagine, finance companies aren't the fastest to integrate with, and in their defense, they have regulation and compliance to handle. And continuing to expand that is something where we've been gaining ground, and we intend to push harder and then extend from there. To address your question, what we aim to do in the nearer term is let those be compared against each other and in one place, ideally with some kind of measure or flag, somewhat similar to the price flags we use, showing which ones could be the best deal for a particular consumer. Then when it comes to the OEMs specifically, there are lots of new OEMs. Every one of them has brands nobody in the U.K. recognises. I'd say they're all highly pragmatic. So they're looking — they view Auto Trader as somewhere they can become known for cars quickly. So ensuring we have the relationships and every solution that lets them do as much as they possibly can or want to do on Auto Trader is part of what underpins the stronger M&A figure Jamie mentioned earlier. And there's a chance for us to simply do more than that — more of that, sorry; and after that it comes down to leads and lead generation, and that's an area where it's still fairly nascent, honestly. A lot of OEMs gauge all their effectiveness on inquiries. Turning that activity into concrete leads for an OEM rather than a retailer falls into that third focus area; on the previous question, we're close to running out of time. Do shifts involving AI and consumer behavior impact how well we can attribute sales? Honestly — and I'd be glad to walk any analysts through it who are interested — our attribution has improved dramatically this year, in fact just over the past few months. Jamie brought up matched sales — that's the case where we've provided data on a retailer's sales and can tie it back, and I mean in a very precise way. It's not just telling them a user in this postcode viewed this car; we go much deeper than that and give them confidence that yes, this came from Auto Trader. We'll supply the time, the date, the stance, what action they took, and which other vehicles they considered. From there it receives a propensity score reflecting how likely it is that we influenced that sale. These days retailers don't really push back on us about that. The interesting piece now is that this is all based on match data. Here's what we've done since — and I realize I'm leaning on this AI example more than I'd intended — we now take that behavior and model it, then extend it across every dealer on Auto Trader, regardless of whether their sales data reaches us. That means we can supply a projected sales figure even without your data, at the same granular level of detail, and start that conversation with dealers. So forecasting sales is no longer dependent on getting their data, and it allows us to have the appropriate conversation with a retailer who says the number we show for their sales is wrong. Our response is simply that we will supply match data, along with all the insight we can offer. They can then see insight into the geography of where their cars are being sold, what others are doing, the patterns of behavior, and what they might want to source. This isn't a topic we've discussed publicly, but I'm genuinely very proud of what the team has accomplished in this area. In my view, it ranks among the finest work I've seen from marketplaces anywhere in the world.
Jamie Warner
executiveGreat. Thanks, everyone, for joining this morning.
Nathan Coe
executiveThanks a lot.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Autotrader Group plc transcript — plus 132,000+ transcripts from 7,000+ companies, speaker segments, AI summaries and full-text search — through the EarningsCalls.dev API.
Get the API View API docs →This call discussed
For developers and AI pipelines
Programmatic access to Autotrader Group plc earnings transcripts and 132,000+ others is available through the
EarningsCalls.dev REST API. Plans from $39.99/month — full transcripts, speaker segments,
full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.
More tools for stock research
OptionsBell: an unusual options activity scanner that tracks options flow across all US stocks and emails you when large trades appear in Autotrader Group plc or any ticker on your watchlist.
StockMarketScan: a stock screener with alerts: 24 pre-built screens refreshed every 5 minutes, watchlists and email notifications. Both offer API and MCP access and fair monthly pricing.