Somm.ai lets you drink like a pro. We use data to analyze wines to help you feel more confident when you’re ordering a wine. How can we prove to you our scores are meaningful?

After a month of seeing users, we noticed that there was a big discrepancy in the usage between our users. Somm.ai has a “pro” version which is meant for people who like wine so much that they try to find specific wines or producers.

We looked at what the PRO users were searching for. The differences were vast compared to regular users.

We rated the median producer searched for by our PRO users as 94% vs. 55% for our regular users.

But don’t think that our PRO users just have a greater wallet. The median wine price for a PRO user was $211 vs $185 for our regular user.

That’s not the craziest thing.

The median retail price of our PRO user was $240 vs $71 for our regular user. So our median PRO user was finding wines that were much better in quality (judged by our ratings and as well as the retail price) yet they were paying only 14% more for it!

Our median user was paying a 2.6x markup, which is probably normal.

Our PRO user was getting wines that were selling below retail, which are more likely for the wines in the quality level we are talking about because the retail price is more likely to reflect the market price, whereas restaurants are more likely to price on a cost-plus basis.

Somm.ai stresses our producer score and value - both of which are used in our aggregate score.

You don’t have to subscribe to Somm.ai Pro to get any of these benefits. We sort wines at restaurants based on the metrics we discussed here today.