New article: Pharma’s Data Problem

Marco Schmidt, the CSO of, and I have published an opinion article in the Pharma Boardroom online trade journal. We have been discussing the topic ever since my first keynote talk at the launch of the Charité Digital Health Forum in 2018 (a video of the same talk is available from PyData 2018).

We title the article: Pharma’s Data Problem.

The executive-level motivation behind the article is: why is the introduction of data / AI innovation not having the expected impact in pharma?

To answer the question we coined at least one neologism: Deep Data. I think Wide Data existed already, in another domain, but possibly with a slightly different meaning. In our definitions, Deep Data contains a high number of entries (N) and a low to moderate number of dimensions (D). Wide Data in contrast, has much lower N and relatively higher D.

Big Data comes in one of these two types. The data that Facebook, Google and Netflix rely upon is more accurately classified as Deep Data. In contrast, biomedical data is Wide Data.

One of the major mistakes which pharma managers are making is they are mistaking their wide data for deep data. There is, of course, clear overlap between the domains. But, in general, you cannot apply the deep data techniques, in the same manner in which they operate in the consumer tech industry, to biomedical problems.

That’s it. The message is short. We want non-mathematical executives to understand it. We name-check some potential solutions, but that is really a discussion to be had with the AI/Data leads directly.

Please do read the article and tell me what you think of it. I am keen to have some feedback.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.