The day I knew our marketplace strategy was fundamentally wrong started with a dealer forwarding me a screenshot. Our marketplace connected people selling their used cars (supply) with dealers looking to buy them (demand). Instead of a car exterior, the main advert photo showed someone's sandwich and chips on a desk. 'Is this a joke?' the message read.
It wasn't a joke. It was our strategy working exactly as designed. Upload one photo, go live instantly, no questions asked. The seller had (at least allegedly) grabbed the wrong image from their phone.
That dealer never bid on our platform again. Neither did the three others who saw that listing. In one moment, I realised we'd been optimising for the wrong problem entirely.
Lenny wrote a great article on how to kickstart & scale a marketplace, noting that the vast majority of successful marketplaces focused almost all of their resources on growing supply early on (80% of the companies interviewed). Our marketplace allowed users to list their vehicle to sell (supply) and dealers (demand) would bid on the vehicle in order to buy them. Initially our supply was high (although poor quality) and we had product market fit with a segment of dealers, but we had very little liquidity.
The sandwich photo taught me that optimising for good supply is optimising for demand, just on a longer time horizon. We'd been thinking about supply and demand as separate problems, but that dealer's reaction showed they were the same problem. Every low-quality listing didn't just waste a seller's time; it actively damaged our relationship with our dealers. We weren't just failing to serve demand, we were training them to go elsewhere.
But the real lesson wasn't about photo quality, it was about human psychology. That dealer didn't just see a bad listing; they saw a platform that treated their time and expertise as worthless. We'd missed the human element entirely.
“Liquidity isn’t just a metric. It’s the result of trust — and trust is built in the smallest details: the photo you approve, the timing you set, the predictability you deliver.”
Jack Burrows, Product Director @ Carwow
How We Rebuilt Dealer Trust (And What It Actually Took)
We looked to rebuild trust step by step, mainly through dealer conversations and feedback. What emerged were three fundamental shifts that transformed not just our platform, but how we thought about marketplace dynamics entirely.
1. Finding PMF with the Dealers Who Could Scale Us
We'd found initial PMF with dealers who had time to hunt for bargains, independent operators happy to sift through listings for the occasional gem. But to scale, we needed much bigger dealers who could buy vehicles in bulk and provide real marketplace liquidity.
The challenge was these dealer types had conflicting needs. Our bargain hunters didn't want competition, they wanted to snap up underpriced stock as quickly as possible. But we needed multiple bids per vehicle to get sellers good prices and stop them going elsewhere to complete the sale. That tension ran deeper than just dealer segments, it was built into the marketplace itself. Sellers valued speed, simplicity, and the best price. Dealers valued accuracy, predictability, and margin. Optimising for one side often came at the direct expense of the other. Balancing those motivations wasn’t about finding a middle ground; it was about designing systems that could make both sides feel like they were winning at the same time.
The bigger dealers also made it clear: they wanted us to remove any admin from their workload. They didn't want to spend 20 minutes investigating whether a listing was legitimate. They wanted pre-verified inventory they could evaluate and bid on quickly.
We weren't just serving different customer segments, we were trying to serve fundamentally different business models.
2. Thinking in Systems, Not Features
This changed how we approached every decision. Instead of asking "what feature should we build next?" we started asking "what behaviour are we encouraging?"
When we introduced stricter photo requirements, we weren't just trying to improve listing quality, we were showing dealers that we understood their world. Every decision we made sent a message about what kind of platform we were building.
We'd moved from building features to shaping how our marketplace actually worked.
Through countless dealer conversations, one phrase kept coming up: "We don't like surprises." A sandwich photo was not just embarrassing, it was unpredictable. Dealers needed to know that when they spent time evaluating a listing, they would not be walking into an unknown. Their business depended on reliability, and we had built a platform around uncertainty.
We shipped a guided photo experience to introduce consistency in advert images, reviewed every listing for accuracy before it went live, moved the valuation from the end of the flow to the beginning, and added friction to encourage users to upload photos of damage and their service history record. These changes signalled care, attention, and a commitment to reliability, which in turn rebuilt trust with our dealers.
3. Obsessing Over Our Liquidity Metric
While our north star was always sales, listings with 3+ offers became our true liquidity metric because it captured both sides of the equation: sellers getting competitive offers and dealers finding inventory worth fighting for. Real liquidity wasn't about having 1,000s of listings - it was about having listings that dealers actually competed for.
But tracking the metric was just the beginning—optimising for it changed everything about how our marketplace operated.
The moment we started obsessing over our 3+ offer rate, it became clear that our biggest problem wasn't supply quality alone, it was timing and urgency. We had dealers browsing sporadically and stock sitting static for 7 days. No wonder we weren't creating competitive bidding.
The solution emerged from our metric: we launched a daily auction format. Instead of listings going live randomly and staying there for a week, we batched approved inventory into daily auctions with clear bidding windows. This created two crucial behaviors: dealers developed a habit of logging on every day, and they knew they had a limited timeframe to place bids.
We also tried to reduce friction by moving from double-commit to single-commit. By collecting reserve prices upfront from sellers, we could automatically award vehicles to the highest bidder above reserve, eliminating the back-and-forth negotiation that often killed deals.
The Payoff of Trust at Scale
The hardest part wasn’t adding features. It was earning trust at scale. Every high-quality listing signalled to dealers: we value your time, your bids, your business.
Liquidity isn’t just a metric. It’s the result of trust, and trust is built in the smallest details: the photo you approve, the timing you set, the predictability you deliver.
The day I saw our first daily auction with multiple competitive bids on every listing, I thought back to the sandwich photo. That dealer had walked away because they could not trust what they would find when they clicked. Now, dealers were showing up every day, knowing exactly what to expect and eager to compete. That is when I knew we had built trust at scale.
ABOUT THE AUTHOR
Jack has spent the past eight years at Carwow, helping orchestrate the company’s evolution from startup to scale-up. He launched and scaled Sell My Car into a core business line, built carwow’s first machine learning team to embed AI into the core experience, and led product across leasing, marketing, and used cars.
Energised by the challenge of scaling marketplaces, Jack’s focus is on marrying commercial impact with user trust. He thrives on building high-trust, high-performing teams, and is equally at home shaping long-term strategy as he is driving day-to-day execution.
