Well, yes. We really like the Mom Test here, as you may have picked up….
I used the Mom Test this week when talking to a Professor of Artificial Intelligence at Edinburgh University called Bill. I was doing some DD into an NLP & machine learning startup and needed to get his honest view on the business after he had had a 2 hour call with them on our behalf (and our dime).
Something the Mom Test encourages is to ask important questions. What’s an important question? If you get an unexpected answer to the question and it doesn’t impact your plans, it wasn’t an important question. Taking a step further, every time you talk to someone you should ask at least one question that could totally destroy your currently imagined business. Now THAT is an important question.
So back to Bill. We originally had a list of questions like this:
- How is language translation done?
- Did they show you the quality of output and what is the proportion of sources between news / blog / twitter / other social media etc?
- Did they show a live demo on the feed analysis from different sources?
- How much time is required to create an analysis?
They’re fine questions, and good to ask.
But after discussing with the deal team for an hour – yes, we prepared our questions for a whole hour – we came up with some better questions. Some important ones. Here are some examples:
- As a professor of AI, does anything genuinely impress you about [company X]? Would anything they are building place them at the top of your class?
[I thought this would help him contextualise the team in a way with which he was familiar – his life as a Professor. It worked!]
- How would you have built the system they have built…….
- …….How different is [company X]?
[It was important to separate these 2 questions so he wasn’t biased in his response to the 1st]
- Their techniques – bleeding edge, leading edge or last year’s thing?
[It was very important to us to know that these guys are at the cutting edge. This was a similar question to the first, but we wanted double reassurance and to see if he was consisent in his answers]
- What are the key areas where they need to focus in the 6 months from an AI and/or architecture perspective?
The answers to questions like these got us so much more than the answers to the first set of questions. The answers would genuinely get us closer to an invest/don’t invest conclusion whereas the former list of questions would just nibble around the edges of that question.
Some other advice from the Mom Test that I love is:
- Have they even googled a solution to the problem you are solving? If not, it’s not a problem
- How much does the problem cost them? They may get emotional about problems that don’t actually cost them much. You have to see beyond the emotion
- Love bad news – gets you closer to the truth
- Lukewarm is a great response – you know the person doesn’t care
- Pre-plan the 3 most important things you need to get out of the interview and ensure you get them answered
- Commitment and advancement: you want them to give up something of value – money, time, high value connections (reputation risk) to prove they care about the problem and you. Make sure they shift from fuzzy future promises to concrete current commitments
- By taking an interest in the problems & minutia of their day you’re already more interesting than 99% of the people they’ve ever met. Just ask about their life
This is just a taster of lots of goodies in the book. Go Mom!