The end and the beginning

In February 2021 Auto1 completed a successful IPO with co-founders Hakan Koc and Christian Bertermann at the helm of the company. Ultimately the Auto1 story is for the founders to tell but perhaps my perspective will be of interest until then. In particular what always impressed me about the team is that they always worked hard to get the details right without losing focus on their strategic goals. That is one key lesson I believe others would do well to learn from.

I was introduced to Hakan by my former colleague Daniel Jones in 2013. The company was very close to agreeing a deal with some other parties so my first challenge was to persuade Hakan we should even meet. I vividly remember standing in the rain outside a restaurant where we were meeting a candidate for our team and persuading Hakan to at least take a meeting before he signed. By a lucky coincidence, his mother was from the same town as my wife (Edirne, the former Ottoman capital) and that was enough to get the first meeting.

A €350bn strategic goal

From that first meeting to IPO the core thesis didn’t change, transform the €350bn European used car market by making it easy for consumers to get a fair price when selling used cars and for dealers to easily access inventory that consumers wanted to buy. The European used car market is €350bn, highly fragmented and limited in its use of technology. Most consumers and businesses sell their vehicles either privately in person, at a dealership or via offline auctions and all rely on an in-person vehicle inspection before buying. What got me excited was the opportunity to create a used car marketplace where Auto1 was the exchange, a market maker and a clearing house all rolled into one.

In public equity markets exchanges standardise terms for transactions and create a venue for buyers and sellers to meet. Market makers help bridge the gaps between buyers and sellers and therefore bring liquidity to the markets on behalf of the exchange. Finally, given the crowd of buyers and sellers on the market payments are handled by clearing and settlement systems allowing complete strangers to trust their transaction will be completed. No one these days would advocate for the old days of the offline stockbroker who established prices by chatting with their friends and whom you could never be entirely sure was giving you a fair deal but that is how the used car market still operates in most of the world.

For consumers, their car is their most valuable possession after a house so any transaction is stressful. Sellers are never quite sure they got a fair price and most spend months dealing with potential buyers online and offline before they manage to sell. For buyers the used car market is the most famous example of a market for lemons where buyers should beware and not much has changed in the almost 50 years since Akerlof published his article. This was the problem that Auto1 was looking to resolve for both sides.

The first three branches

We first invested in Auto1 when they had three branches in Hamburg, Cologne and Berlin – the latter was also their head office. There was a covered garage with a wirkaufendeinauto (“we buy your car”) banner hanging from it, a large unsurfaced parking lot and an office with a dozen rooms coming off a single long corridor. At the beginning most of the rooms were empty but each time I went there was more and more activity.

Auto1 was a classic fast follower as WeBuyAnyCar had proven a model like this could work in the UK but there they sold most of their inventory to British Car Auctions (BCA) rather than directly having relationships with dealers. Auto1 was operating in a market where auctions were not established and therefore had the more ambitious goal of directly building dealer relationships. It required a direct to consumer operation attracting people who wanted to sell their cars, a sophisticated purchasing operation and a sales machine that could attract dealers to buy they cars. All the while the company had to manage the cash burden of carrying the cars on their balance sheet between the purchase and sale. 

One of the challenges of a marketplace for physical products is agreeing what exactly is being bought or sold. As the railroad expanded in the US shared wagons became the most economical way to move grain so farmers would mix their grain during transport and needed to know what quality of grain each contributed. In parallel the Chicago Board of Trade was creating a market for it and the result was the “Grain Standards Act” defining quality tiers and a system of measurement. In the new car market there are already hundreds of variants of the “same” car such as a VW Gold or a Renault Clio based on trim, engine specifications and then all the optional extras. By the time it has been driven for five years you have to factor in the mileage, service history, rusting, dents, scratches, interior cleanliness and how much it smells of cigarettes or dogs.

Auto1’s goal with dealers was to do a better job of inspecting the car than your typical dealer would be able to so they would buy it online. You had to capture every detail of the car being traded in order for buyers to trust you enough not to want to come and see the car for themselves. In practice this meant their condition reports would be setting a standard for the industry. At the beginning the pricing process was largely manual, there was some automated scraping of public data but Christian and his pricing team decided on the bids for each car. Over time as they captured the data to understand what influenced the price of a car they started to automate this more and more. 

From the beginning the team obsessed over data that could help them improve the business and were not afraid to pick up the phone to hear customer feedback directly. For the consumer selling their car price was one component in addition to the next day payment and ease of selling but it also became clear that there were a number of non-tech factors. The purchaser that welcomed the customer and walked them through the sales process, as well as ultimately communicating the offer price, was a key determinant of conversion rates. The team therefore put in place a training program to spread best practice across the organisation. No feedback was considered too trivial, for example when sellers complained about hand towels running out in the bathrooms at branches that went on to the checklist for the branch staff. All of these details added up into steadily improving conversion.

Scaling up and experimenting

As the business expanded from 3 to 15 to 45 branches over the next funding rounds and with revenue scaling in parallel the approaches changed but the focus on understanding the details never did. I recall going through the entire value chain with Hakan and Christian during our due diligence from lead generation through to the sale to the dealer and for every question Christian would write a SQL query to answer me in my favourite language: data. When there was no data readily available the team did not hesitate to go and get it and if there was a job that needed doing they dealt with it.

As a first example, when the team wanted to explore new markets for the very oldest cars they wondered about the trade to Africa. It was known that many of the cars with the highest mileage were traded there so the team sent a shipment and Hakan to test prices and identify the dealers there. We heard from him that in Douala many cars were unloaded that were stuffed full of broken stools, old televisions and other items for trade – the car was just the container. This was not something you could know from afar without going and having seen for yourself what the unloading and the marketplace there looked like. It was the information the team needed to decide it was not a route worth spending time on.

Similarly, when the company tested the US market in Texas they were back to startup mode with a limited team. This was not the usual slow, multi-million dollar experiment in US expansion but the creation of a proper startup team. The first time I called, Hakan was washing windows at the branch, another one of the details that they had learned made a difference to the consumer selling their car.

Don’t get distracted

As companies grow they often get distracted from what let them grow in the first place. Customer success teams get between product teams and the end user, the product gets bloated or management distracted by new projects or M&A. In the first Auto1 investment presentation there was a page describing future growth opportunities that included floorplan financing for dealers, selling cars direct to consumers and monetising their data. For each of the next two funding rounds the slide remained unchanged and these remained on the roadmap. Given how much opportunity was ahead of the team with their core value proposition they focused all of their resources on the proven model that they knew how to grow while those resources were still limited. Don’t get distracted by the next growth opportunity until you are sure you have the resources (management time and money) to push them in parallel.

The team also avoided spending money on anything that didn’t help improve the business. When the business moved HQ from the Pankow branch in Berlin to a ‘proper’ office there was a reception desk at the entrance and we joked about the contrast to the prior office where we had to check each of the rooms on the corridor to find the team. A few months later it became a standing joke that they never hired a receptionist to sit at the desk, it was not one of the things that helped them grow the business faster.

Publicity and fundraising are often also a distraction to teams and far too many companies end up measuring their success by their news articles or capital raised. Auto1 clearly raised a large amount of money in their path to IPO, which was a requirement given they hold cars on balance sheet, but it was never the goal. When we first invested we kept the funding round private so as to avoid inspiring competitors and for a long time those who knew the team thought they were building a fraud product. However, they did extensive marketing and publicity amongst used car dealers to raise their brand profile – thankfully most VCs didn’t read the same magazines! 

Looks easy from the outside

All great achievements can look slightly magical when you view them from the outside. I met one of the management team at BCA early on who was convinced that the Auto1 vehicle listings must be fake. He didn’t believe that they could have grown that fast and really have that many cars for sale. I suggested he try buying them if he had any doubts and from the wry response it was clear they already had but still couldn’t quite believe it. I had the same response from one of my partners during a follow-on round, he wanted me to drive out and check the cars were really at the logistics partner.

When I first invested I took extensive references on the founders and became convinced by their drive, intelligence and determination. If anything, people understated how hard they worked and how much they would drive the people around them. Through everything they never lost sight of their goal to build a public company that would transform the European used car market. What people miss when they read about such successes are the thousands of tiny details needed to deliver such a result. Congratulations again to the founders and all the team at Auto1 for getting them right.