,
7 tweets,
2 min read
Read on Twitter
1/ Thanks @keerthivasanck - I'm assuming @asanwal means tech vc when he says vc. (Most people do) In that case I agree with him. For data to be effective helping drive private investment decisions we think there needs to be a) similar business models (tech doesn't have this)
2/ b) a ton of data- feature data to evaluate the co's and training data upon which to build the models. Neither exist in tech.
3/ Feature data doesn't exist in big ways because hard to get reliable data externally. Hard to get customer list, pricing list, what end users think of the product, etc in big reliable ways. Lots of firms have tried including buying data from providers.
4/ Credit card data for some tech vc applications has been somewhat helpful but hard to go wide, and more useful for coming deep. Tech VCs have tried other options like taking data from Crunchbase, Prequin, scraping LinkedIn and some others.
5/ [Fun fact- the old CSC Upshot fund which I *think* closed down after the founders of AL left AL found that the only signal that had a decent correlation with success was who the third best investor in a round was.....depressing]
6/ c) lack of adverse selection. Even if the data informed VC in tech gets there first, they are still going to go to one of the Top Tier funds.
7/ I dont believe a, b and c are true in all sectors. Clearly I'm biased but here is how I think it would work in consumer with co's $1m-20m in rev
