How a bot could automate 80% of startup investment screening

I have a 100-step checklist that I use to screen new investments. I’ve backed about 70 companies, but I’ve evaluated 40x that many — a dizzying amount of research. The fact is, a bot could do 80 percent of that work for me. After all, I have a series of questions, and I need to fill in answers. This is the type of problem bots excel at solving.

This week, a team at Gigster built a model of what this bot might look like: Pitchbot.vc. It doesn’t ask all 100 of my questions, but it covers much of the same ground that I do as an angel investor. For example, the bot asks questions about who is on the founding team, what the market looks like, and how the idea has been validated.

Entrepreneurs can use Pitchbot to get a sense for the flow of conversation with an investor and can even customize the bot to respond like an incubator, a seed fund, or a high-profile VC firm, depending on whether they want to pitch Y Combinator, First Round Capital, or a16z.

Pitchbot screened 5,000 startup pitches in its first 24 hours of operation. If I were screening companies full-time, I might be able to research around 100.

I’m not about to take the proposed valuation and stake that Pitchbot offers entrepreneurs at the end of the Q&A and run with it, but it could still provide value to me as an investor. For instance, I would prioritize my own research based on its rankings.

It wouldn’t be too hard to build a bot that I trusted to get more reliable results. After all, Pitchbot was built as a tool for entrepreneurs, not for VCs, and yet it’s already taking some work off of my plate.

Democratizing VC time with data-driven choices

Screening startups is a pain point for VCs. They know that they’re letting great opportunities slip through the cracks because there are so many new ideas but only so many hours in the day. They have to rely on tips from insiders and other trusted parties to make investment decisions. At the same time, they are committed to advising and helping their portfolio companies, where they can provide startups with valuable direction and connections.

Bots could give venture capitalists the best of both worlds: they can screen startups systematically while redistributing the VC’s time to portfolio companies and making high-level decisions. There will always be a gut factor in investing — to trust the founders, to see the potential for a new market, or to catch a mistaken hypothesis — but by using bots to do the first round of research, VCs would have time to give these high-level questions more thought.

It’s only fair that VC gets automated first

Silicon Valley venture capitalists are all bullish on automation as a piece of the future of work. Investors pour billions of dollars into autonomous driving, workflow tools, VR, and other technologies that will make many jobs almost unrecognizable in the next 10 years.

Why is this a priority? Augmentation and automation can improve everyone’s quality of life and create new avenues for economic growth. In the case of VC, a pitch-assessment bot could do both very quickly. The VC will be able to focus on problems that are most likely to give portfolio companies a better chance at the big time, and budding entrepreneurs are promised an equal chance at getting noticed, even if they don’t have the connections for the fast track to Sand Hill Road.

The future of work is right around the corner, and the future of the economy rests in the hands of the country’s entrepreneurs, as well as its international enterprises. If large parts of our work are going to be automated, we need to keep lifting up the best ideas to generate new opportunities. If we let bots take over the time-sucks, then we’ll really see what people can do.