Imagine walking into a used-car showroom and driving away in a newly purchased vehicle—without ever encountering a live salesperson. That’s one way to buy a used car from Carro, an online marketplace for automotive and related services founded in Singapore in 2015 by Aaron Tan. Since its founding, Carro has enjoyed impressive growth under Tan’s leadership. McKinsey’s Mudasar Mohamed recently sat down with Tan to discuss Carro’s growth trajectory and value proposition, as well as the role of AI in supporting productive and profitable growth. The following is an edited version of their conversation.
Key insight #1: Moderate your growth, set achievable targets, and hit them.
Mudasar Mohamed: To what do you attribute your high growth and profitability in recent years?
Aaron Tan: First, a lot of companies make the mistake of growing too quickly. The faster you grow, the more likely you are to make errors that could result in losses. It’s important to moderate your growth. For a series A, B, or C company, 200 or 300 percent year-on-year growth is enough; anything more than that is too fast. We have a growth plan, and we review it every year.
Second, it’s important to hit the projections you shared with your investors, or you will quickly lose their trust. A lot of companies make the mistake of going out with very aggressive numbers, missing them, and assuming investors will forget. But venture funds hold you to your targets.
Mudasar Mohamed: With respect to geographic expansion, what criteria inform your decisions about which new markets to enter?
Aaron Tan: First, part of not growing the business too quickly is not expanding into new markets prematurely. We really work to understand a market before we enter it. Every market is a little bit different. We cannot assume that what worked well in Malaysia will work in Indonesia or that what worked in Singapore will work in Japan. In Japan, for example, the market is very mature; the supply of used vehicles is high, and USS already operates the country’s largest used-vehicle auction marketplace, so we opted not to launch auctions there. With these differences in mind, we have built our systems, processes, and teams to be very modular so that we can adapt as we expand.
In addition to sizing a market, we also think about a market’s readiness by using metrics such as rental rates, unit sales potential, and more. We also look at the areas we want to compete in. And nowadays, we concentrate on trying to determine when the market is going to break even.
In traditional ventures, there’s a lot of focus on the J-curve. Over the past 12 months or so, however, we’ve started to think internally about how to flatten the J-curve so that we burn less cash initially and grow more gradually. This also ties into our expansion strategy.
Now that we are getting bigger as a company, timing the market is very important.
Key insight #2: Whether your company is physical, digital, or both, it’s the customer experience that matters.
Mudasar Mohamed: Does Carro’s business lend itself to a blended online and offline model?
Aaron Tan: We focus on the supply side of the auto business, and we source cars from online and offline channels. Our motto is “as good as new,” so selling peace of mind to our customers is central to our value proposition. It’s similar to the concept of “certified pre-owned” in the United States. We believe that a good car sells itself, so we don’t worry too much about marketing, whether online or offline; we’re more concerned with delivering a differentiated experience. Ultimately, of course, the experience is “phygital” because a customer might start their journey online, but we deliver a physical vehicle.
Customers can visit our showrooms, or what we call “pods,” at any time, 24/7. They can complete the end-to-end car-buying journey 100 percent online. Customers tell us they appreciate the efficiency and report having a better overall car-buying experience. Customers also trust that they can come to us with any problems with their car because we also run our own repair shops.
Mudasar Mohamed: What are some of the performance metrics that you, as CEO, track most closely at this stage of Carro’s growth?
Aaron Tan: The three metrics I’m always tracking are revenues, EBITDA, and cash flow because those are the ones our investors care most about. But we have more than 500 KPIs that we track using dashboards. We rely a lot more on data than we did two or three years ago and have really strengthened our data capabilities. It’s very important to have an internal data culture. It starts from the top, but everyone needs to understand the power of data.
Key insight #3: AI is a powerful productivity tool that supports profitable growth.
Mudasar Mohamed: Can you explain the role of technology, especially AI, at Carro?
Aaron Tan: We believe in the power of data and AI. AI boosts productivity and can help reduce costs, increase revenues, and boost earnings. Consider car inspections. Inspections by humans can be highly variable, even for the same car, depending on the inspector’s mood, level of expertise, and so on. Inspector bribery is also a very real concern in some of the emerging markets where we operate. In these ways, humans are unreliable.
An AI-led process, on the other hand, is devoid of emotion. Inputs and outputs are delivered in a structured way. We input pictures of vehicles and use AI to detect scratches and even accident history. We use this assessment to determine a rough estimate of repair costs and make informed decisions about which vehicles to buy. The AI system is continuously training itself with every car it assesses. Humans are much more likely to miss problems, and those omissions can translate to costs. We also developed an AI system to assess the car’s internal condition using sound as an input. Sounds of an engine misfiring, for example, could signify a problem such as the car having been flooded in the past.
Mudasar Mohamed: How has your leadership style evolved across the various stages of growth of the company?
Aaron Tan: I have always believed in leading by doing. I still code a lot. I always believe that if I’m running the business correctly, then I’m delegating, which means my time and attention aren’t required at every turn. Having said that, my level of involvement varies based on the situation. I will micromanage to make sure we close a round of financing on time, but I’m rarely involved in day-to-day operations.