How do you measure the success of your data (and data science) products? Any reading on the topic you can recommend?
(Seems to be sparse content on the topic online. 
)
Show this thread
-
@imightbemary that's a good point. Can see how a support team may have different deliverables (and thus measureables) than a research team. In this case, it is an R&D team building ad targeting products.
You’re fundamentally looking for a counterfactual. What would the state of the world be if you didn’t have this team or data product? Controlled trials (A/B tests) can be helpful for this, as can casual inference models or natural experiments.