AI startup investment criteria

We’re all investing time or money — let’s make sure not to waste it

I started to collect a lot of criteria I thought was a good investment of my time and had a good chance of success based on experiences I’ve had creating a lot of products in the past (like social search, IoT, logistics, AI, and blockchain).

For this post I’m just focusing on 1) Applied AI startups that 2) want to make money reasonably soon. I will not include in this discussion: 1) Research AI startups that 2) get tens of millions (I would want at least $50MM) in order to spend a decade doing tech-driven R&D in hopes of a product (or acquisition) in the future. Nothing wrong with that. It’s just a different category.

Caveats:

Most of these criteria are useful for all startups, a few are particular to AI startups.

These are general thoughts/ guidelines — there will be counter examples/ anecdotes of startups that break most or all of these rules. They are however, useful guidelines to think through. And, you will want to customize them for your own situation.

AI Investment Criteria for Investors and Entrepreneurs

Is it a learning product? (product gets smarter with usage)
With an AI product, you should see an AI flywheel impact. What information are you collecting and how is it making your product smarter?

Where are you getting your data?
Look, there are a lot of AI ideas out there. If you are a startup, I would tread lightly into territories where AI is using public data or easily collectable data (that large companies already have). As an AI startup, you need to collect data from a closed system that you can create proprietary insights around. This may mean that your data and training is something to work out between you and your customers. If it’s unique enough, they will want to come to you to provide insights and features they can’t get from the big companies.

Does it provide custom value per customer?
Creating something that creates custom value for each customer is a more sticky value proposition.