How did Tractable get started?

We are an applied AI company and we primarily specialize in transforming the automotive and insurance industries. Our core focus is on developing AI-driven solutions that streamline and enhance vehicle condition assessments. Our initial use cases and customers were primarily in the collision repair, insurance, and auto recycling industries. We can look at a picture of a car and identify the parts that are damaged, the length, breadth, width, and severity of the damage. And depending on the use case, we can recommend whether to repair or replace the car. Working with customer-specific inputs and data from our partners, we can come up with an estimate of costs very quickly. For a car in the recycling industry, we can determine which parts are reusable.

Things have changed dramatically from a decade ago, when this kind of technology seemed like magic. How has that impacted Tractable?

Yes. The bar is very, very high. We have to show real value, especially as a company in the B2B space. If you look at the creative and consumer space, when things go wrong with AI, the threshold for error is fairly large. ChatGPT can write you an email, and if it’s not right, you tweak it and send it. It’s OK if it’s 80% right. We live in a vertical AI world. We talk about the assessment on a car that will lead to actual money being exchanged. There’s a difference if a car is a total loss or not; if it goes to a junkyard or gets repaired. Those are high-value decisions. So we have to focus on the last 20% in accuracy. We have to be able to solve the business case with a high degree of confidence. If we go too far out and say, “I’m going to make magic,” we risk coming up short and losing everything. If we go too lightweight, it waters down our value proposition. So we have to find that Goldilocks zone of use cases that match our capabilities and deliver strong outcomes. It has changed how we approach the market.

How do you tackle that last 20% problem?

Every customer, insurer, body shop, and recycler we work with has their own secret sauce. We built our product with computer vision tech at its core, and with a very deep and scalable customization layer. Some of our customers have enormous amounts of data, for example, on how they buy cars for recycling. With their proprietary data and logic and our core AI, we are able to make something that’s very unique to them. For that last 20%, we continuously push the bar on how good it is. But the full solution can’t be done without the customer’s secret sauce.

There’s a huge amount of FOMO in the market around AI, both among customers who are buying tech, and among companies who are making tech. There’s this idea that if you’re not on board with AI, you’ll be left behind. How is that playing in your market?

I think the FOMO is there, but there is an equal amount of skepticism that’s emerged recently. We operate in a highly regulated industry across markets in North America, Europe, and Japan. And, when it comes to the work that we do, which is the actual claim processing, actual repairs of the cars and such, I don’t see FOMO as much as a more cautious approach from customers. I strongly believe the market is still in its early stages. In fact, Tractable has been a little ahead of the market in some cases. We need to grow with the market, and bring customers along. So, I don’t think FOMO exists with our customers. They have a desire to embrace AI, but they want to do it the right way because of regulatory compliance needs.

Is there a lesson for other entrepreneurs in that?

Don’t try to boil the ocean. Don’t get ahead of the customers. That’s the biggest mistake people make. People’s perceptions need to change. Confidence needs to be built. Business processes need to change. It’s a big lesson in how to think about solving small problems while having big aspirations and ambitions. You can’t jump 10 steps ahead of where the customers are.

Does this relate back to this idea of “outcome, outcome, outcome”?

Tech is useless without business impact. And the fastest way for a customer to have business impact is to apply technology that delivers additional value in areas that are already working. It’s why we don’t chase revenue anymore. We are chasing outcome and value, and we strongly believe that revenue follows very quickly.

What lessons have you learned in the process of leading Tractable through this change?

As tech people, we dream up scenarios of what the industry can be and what is possible. But we shouldn’t make the customer climb up the mountain. We should go down and meet them where they are and help them in the journey. And that’s a fundamental lesson for me. It’s easier said than done. As technologists, we are all optimistic and believe, “Hey, I could transform this industry overnight.” But you have to meet the customers where they are, make it easy for them, remove all the barriers, and then things will move fast.

That seems a bit counter to the swing-for-the-fences mentality that many investors want to see in entrepreneurs.

I don’t think it means giving up on the big dream. We have a big vision. We are a venture-funded company because we’re willing to take on a challenge no one else is. That’s the reason why we exist, right? I’m not saying, play it safe. I say, play it smart. Because what happens is when you get too far too quickly, you burn more money than you need to. You risk losing all the cash before the market is ready. You might end up with a great product but miss out on the opportunity to do something magical. That’s especially true for B2B companies operating in complex business workflows where trust is paramount.

English

Q&A

Venkat Sathyamurthy: A mantra for AI success

Why “outcome, outcome, outcome” became an obsession for Tractable’s CEO

I strongly believe AI is a team sport and we want to play well with the ecosystem.VENKAT SATHYAMURTHY, CEO, TRACTABLE

You alluded to having changed your go-to-market approach. How and why?

We focused on insurers and repairers early on because that’s where we saw traction. But we have come to realize that the applicability of our tech is much broader: OEMs, used car sales, fleet management and inspection, and so on. The possibilities are endless.

We needed to explore a variety of use cases in a very scalable way, as we believe there is a large surface area of problems we can solve. Our core product is our suite of AI capabilities with a couple of vertical solutions on top. But our platform is our real asset, and we are going partnership-first. We want to embed our AI into as many workflows with our partners to deliver value to our customers. Our aspiration is not to say we build the whole thing end-to-end. Instead we will take our tech and help industry partners embed it and make them better at what they do. I strongly believe AI is a team sport, and we want to play well with the ecosystem.

We believe in our tech so much that we feel it should be everywhere. So, we are trying to remove every constraint, red tape, and technical barrier to get our tech into the market as widely as possible.

Sōzō Pulse

Amid growing optimism, CFOs are stepping up investment and eyeing growth

Explore the full results of our Q2 2024 CFO survey. Compare and benchmark your company

Feature

Job&Talent: Hacking global expansion through M&A

What the staffing marketplace Job&Talent learned about going overseas