Apply Now
Episode 3: The Billion Dollar Story of Tractable

Alex Dalyac is co-founder and CEO of Tractable, which builds “AI for disaster recovery.” Tractable enables auto insurers to make real-time appraisals of damage using their computer vision models. They recently announced their Series D, led by Insight, which values the company at a billion dollars. This makes them Entrepreneur First’s first unicorn and the first unicorn built through a talent investing platform (Alex met his co-founder Razvan at EF). 

In this podcast episode, Alex and Reid Hoffman, co-founder and Chairman of LinkedIn speak entrepreneur to entrepreneur to reflect on their own journeys as unicorn founders. They share what they believe are the most important tenets of founding a successful business – particularly in the earliest stages – and how aspiring founders can launch a startup heading for unicorn status.

Show notes:

Alex explains what Tractable does [1:35]: Briefly explained, it’s computer vision for accident and disaster recovery. In a car accident, fixing the car takes 8 hours, but the damage assessment and claims process can take up to 20 days, which is a huge deadweight loss, which computer vision can dramatically accelerate. Another example is natural disasters where wildfires leave people without a roof over their heads for months. Computer vision can help accelerate that.

Reid’s journey into entrepreneurship stemmed from his interest in people [3:30]. He realised that entrepreneurship would actually be the higher scale impact because it was a kind of software medium that we create in this world around us, allowing us to express our identities, communicate with each other, find other people to collaborate, form and strengthen relationships. He was keen to know how we could help ourselves as individuals and groups make each other better, evolve in capabilities and wisdom and compassion. Alex’s journey in entrepreneurship started with an internship with blitzscaler, Rocket Internet, and he was convinced that he wanted to build a company. 

Reid shares about why technology is at the heart of building huge companies [5:50]: Part of what it is to be human evolves as we evolve our technology, and there are relatively few paths by which an individual can accomplish at a very massive scale. And within blitzscaling, this is a key part of how technology companies of the future are being built [6:15]. 

Note: Blitzscaling is a specific set of practices for igniting and managing dizzying growth; an accelerated path to the stage in a startup’s life-cycle where the most value is created. It prioritizes speed over efficiency in an environment of uncertainty, and allows a company to go from “startup” to “scaleup” at a furious pace that captures the market. For more information, visit

Alex talks about how his experience at Entrepreneur First helped him understand how to make use of technology to build his business: One thing Entrepreneur First taught him was to start with working on a tech problem that might be tied to a breakthrough [8:00]. In 2012, there was a huge breakthrough in deep learning, where it was the first time that they could take imagenet data sets numbered in the millions, feed it to deep learning algorithms and run it on reprogrammed GPUs. With the compute power 100 times stronger than CPUs, this led to a massive breakthrough, where all of a sudden, for the first time, the algorithm started to be able to understand what’s on an image [8:40]. This had implications for real world image classification tasks, where algorithms could take the place of humans. 

Tractable’s strategy to scale revenue was scaling contract sizes among enterprise customers: Most companies think of scaling revenue from 10,000 to a million customers [9:35]. Tractable was focused on getting enterprise customers and the key was scaling from a six figure to eight figure contract. This was evident by helping large insurance companies generate 8 figures in value, which was significant.

LinkedIn’s growth strategy: Reid observed that small differences in a go-to-market strategy can really be an amplifier [10:35]. At Tractable, they were able to start applying this technology and show they have actually reasonably good error bars. This encouraged adoption as it became an economic downside with clear evidence of limited downside in a mis-categorised case is relatively limited, just an economic analysis [10:55]. At LinkedIn, Reid recognised that first mover was not what matters; first mover to scale was, and he shared this in the LinkedIn Series B deck (link) [12:00].

Tractable’s ability to scale is evidenced in its customer trust: In Japan, they currently have all the top five insurers as customers, they can start pooling their historical data and amass training data faster. This would not have been possible without the scale they now have by showing how they have grown from a scrappy startup to a fully automated AI system on a vehicle side [13:30]. 

Choosing the right people to work for your company is a critical part of any entrepreneurial journey: As a young founder, Alex had to work against his bias that it would be hard to convince someone more exceptional and experienced to work for him [14:30], but his third co-founder, Adrian, who was an experienced business executive, gave him the confidence to keep maintaining that high bar. For Reid, he believes that having multiple co-founders is generally better because of the diversity of skill sets and capabilities that enhance the group overall [15:40]. Appealing to talent requires understanding that everyone wants to be a hero in the things that they’re doing. By enabling a path where talent can be a hero by bringing in something the existing team does not have [16:30], the team can succeed all together because they buy into the mission and vision. 

Network effects played a pivotal part in staying ahead of competition: The insurance industry is complex with repairers, insurers and individual customers. By being able to secure the biggest insurers, Tractable was able to get repairers in their network to adopt their product, which in turn made it easier for the next insurer and so on [17:20]. 

Tractable’s fundraising journey wasn’t easy: While Tractable’s timing was incredible, raising capital was hard as they didn’t own the proprietary groundbreaking technology [18:10]. They also did not have a $100 billion TAM (total addressable market) as they focused on building a vertical, full stack solution. If they had gone horizontal and built a platform that interconnected people who are owners of the cars and homes, their insurance companies, or the government agencies like FEMA (Federal Emergency Management Agency) in the US, this would be a very valuable platform [19:00]. That’s a material company. 

Their vision of insurance is to protect every facet of the economy: Tractable’s AI now is one where a user can pull out their phone, scan their car, and Tractable can understand the damage to the car. This information is also important when selling the car, insuring the car and even refinancing its loan. We are getting closer to the day where the visual inspector is in your pocket, with each of these industries worth trillions of dollars [20:30]. 

Reid believes that the best investors provide more than just capital, startups need to look for the contrarian and right VC: The average top tier VC looks at somewhere between 600-800 companies per year and says yes to zero to two of them [21:00]. Lots of times he looks to the contrary because founders don’t need everyone to say yes, just the right VC to say yes and commit to the journey with the team. At Greylock, he picked Airbnb for its Series A despite other members believing it would fail [21:36]. Six months later, they realised this was the kind of investment they should be doing, but Reid saw past the risks because he saw the potential to redefine the industry, which is exactly what Airbnb managed to do [22:00]. 

Reid’s advice to entrepreneurs: Successful entrepreneurs are infinite learners [23:00]. They ask themselves, “How do I do this better?”, “What I learned here?”, “How do I improve it?” For Alex, he pushed on despite being rejected from EF for the first time. He kept building and weaving the path early [23:50]. And when he built Tractable, he had highlighted a set of considerations in driving adoption, go to market, hiring talent – the process of blitzscaling. 

Tech startups will always have an opportunity to compete with some of the biggest tech companies in the world: Every organisation can only have a short number of high priorities, regardless of their size [25:06]. If they’re not prioritised within the organisation, a startup can potentially move faster, be more disruptive and bring in new customers and allies. Reid observed that Tractable was highly focused in its mission despite big companies also doing intense classifiers and deep learning [26:20]. The difference was that none of them were applying it to what Tractable was doing, and with the huge industry, this becomes a defensible and enduring advantage. Network effects are a form of defensibility [26:30]. This can range from owning key channels, key customers to moving much, much faster, and getting that learning curve, ahead of your competitors.

How does a B2B founder from a technical background build a network: Alex believes that advisors can make a huge difference [28:48]. But it’s also a gamble as you’re giving up equity, so he advises that founders need to choose the adviser who has a really high bias to action, and a high tolerance for risk [29:20]. He advises testing them to make introductions, and if they are, then they could be right for the company. Reid concurs [30:34]. Network effects make for network solutions, and being helpful is not just on a product level, but also includes building bridges in terms of surprising connections which can lead to broader networks.