Philippe Laval, chief technology officer at Jolt Capital, discusses its proprietary software platform Ninja, and reflects on the highs and lows of bringing the system to life.

RD: How does your firm incorporate technology into the deal origination process? Are your systems off-the-shelf or built in-house?

Philippe Laval: We have developed our own software tools in our own system to find interesting investment opportunities for Jolt. Every day, partners use this system – which we call Ninja – to get suggestions of companies they could reach out to, or to find out more about a company that they have heard about elsewhere – for example, who its competitors are.

Ninja also indexes all our office emails, so we can share all our knowledge and insights firmwide in an easy and secure way.

While Ninja is our own system, it is like an umbrella. Every time we find a new service that could be useful – for example, the new open AI tools that are emerging – we integrate them into Ninja. 

RD: When did you start implementing Ninja, and what drove you to do so? 

Laval: We started developing Ninja in 2017 and had two key reasons for developing the system. Firstly, we wanted to know how deep the market in growing technology companies was in Europe, and whether it was deep enough that targeting only companies in this category could be a viable investment strategy for us; Ninja gave us a way to conduct a census.

Second, it was the way to scale our business. We felt that there were two ways to attempt to grow. One would be to hire a large amount of junior staff members, send them to a lot of commercial shows and pitch meetings, and have them report back to us. The alternative would be to use technology to give partners direct access to the markets in which they were interested in investing. 

We said from the beginning that we would take the latter option because we did not want too many layers of people between a potential portfolio company and those who make investment decisions.

Cost was not a motivation initially. To create Ninja, we needed to hire a full-time chief technology officer and software developers, and buy data, so at the beginning it was a significant financial investment. Now that we have scaled the project however, cost has become a third reason for having Ninja. The system has become cheaper than having people doing an analogous job would be. 

In addition, with Ninja we have built value inside Jolt and created something that we can continue to use in the future, whereas human knowledge and skills can be lost when people move on. As entrepreneurs, we wanted to build a real company, not just an investment vehicle, and Ninja has become part of that.

RD: What were the biggest challenges you faced when you first implemented Ninja, and what are your current challenges?

Laval: The first challenge at the beginning was understanding what kind of analytical system could be built. The European geographical area we were looking at is very fragmented, perhaps 20 countries speaking many different languages and each with their own different national register for companies. 

Then there was the question of what makes a good investment for us. Can we identify a fixed set of characteristics for a business that Jolt should invest in? The answer is no. 

Next was the challenge of adoption. Every partner was committed to using the system but it was so different to the tools that they were used to using. To be honest, at that stage Ninja was still developing but it had to be used otherwise it would never be a great system. Everyone was very supportive, so in the end it worked.

Now that the system is working better, the challenges are different. The data problem is solved. We now have access to information about almost four million companies and know how to get the information we need about them. Adoption is almost completely solved. People at the firm are so used to Ninja now that it is almost as much a part of their daily working life as email. We were also lucky that our managing partner said nobody should look at non-Ninja companies.

After six years of using Ninja, we have a huge corpus of companies – almost 30,000 – that have been tagged as ‘interesting’ or ‘not interesting’ by every investment partner, so we can use our deep-learning system to tell us how aligned any other company is with Jolt and our goals.

Now we have the challenge of continuing to innovate, in particular to bring something even more useful to the partners: analysis at an ecosystem level. That could mean taking the names of, say, two companies in an industry and having the system build a competition map around them. If you were looking at companies working with drones used in agricultural technology for example, the system might break this field down into subsectors, such as precision agriculture, drone construction, sustainability for drones, and so on, and then show you companies in each one. This is what we are striving for but it is difficult.

The other ongoing challenge is finding good software developers. It is not that easy to get good potential employees to work for an investment firm because they would often rather be at a technology giant like Google or
a startup. 

Then there is the problem of scaling what we already have. Analysing millions of companies is not the same as dealing with 200,000, and we want to keep the system fast and not too expensive. Finally, we would like to move up another level and analyse other GPs as well as companies’ to find out who invests where and how much.

RD: What are some of the successes you have had since starting to use Ninja? What pain points have you been able to eliminate or mitigate?

Laval: When Ninja started and was suggesting a company to a partner, they would decline 80% of the time. Now, it is the opposite. I would say that now about two-thirds of our dealflow is generated by Ninja, and our last four investments all came from the system.

The system will suggest a company to an investor, who will then complete some pre-deal diligence, using the system as a single point of truth to share information. Months will then pass between this first suggestion and an investment, and it becomes a question of human interaction with the founding entrepreneur as we build a common investment case, so the process is the best of both worlds. 

The idea is to give a superpower to the investor at Jolt so they can be more efficient at doing their real job, which is to connect with an entrepreneur, understand their business model and work towards an investment.

RD: In what ways does your use of Ninja extend to the value creation process?

Laval: We use Ninja to look for bolt-on acquisitions, and a little to assess exit potential. We also give our portfolio company executives access to the system, so that they can search for bolt-ons or competitive intelligence themselves.

RD: Do you have any highlights that stand out in that value creation area when it comes to technology?

Laval: I can think of two of our portfolio companies, both of which were suggested to us by Ninja a few years ago, that are now using Ninja to find acquisitions. One of them, a biotech company, is doing as many as two bolt-ons a year.

RD: Do you think LPs see the use of technology during origination and value creation as a differentiating factor for a private equity firm? 

Laval: We have just closed a €100m extension to our fourth fund and we are already starting to raise fund five. Every meeting with an LP or potential LP included a Ninja demonstration. Most of the time they went on to ask for access, which we are happy to grant. 

Ninja also features a lot in our LP marketing in other ways. Every week we publish data, a business map, or something else potentially useful to our LP network that we have drawn directly from the system.