[noplsbecivil]

Yes i did work at one of the above. The car count data was actually more interesting than people even realized. One thing even many quant funds missed was that the quality of our 'raw counts' still depended on cloud cover which was meta data that was for whatever reason something we didn't share (no one asked for it), so the quality of the signal depended on the geography. Ie sunny places had more signal - there was even a short window back in 2016 before credit card data was more heavily used, that you could predict the earnings results of retail chains in socal pretty well with our data.

To your article's point, I believe my company ended up trying to switch to a platform that allowed people to create their own insights, but im not sure how successful that was in the end