I did some assignments for what I believe was ACNielsen when I was in college. I'd get paid about €10 an hour to stand outside an Aldi store asking every customer that came out for their receipt. (Aldi also doesn't share its sales figures.) I'd also have to note every customer that didn't give their receipt. This gave ACNielsen information on the sales for that particular store. They, of course, did this at multiple locations and sold the gathered data to interested parties.

> What PII is usually in them?

It really depends. There are a lot of dynamics at play, and depending on the source of the transaction data it may or may not have identity data tied to it, or be in a format that would easily be identifiable when combined with other data sources. Some sources would consider identity data itself as a competitive advantage, and only sell anonymous or aggregate data to others. Other sources would have no value out of withholding the identity data, and would capitalize on the opportunity to include it in their feed for a premium.

Feel free to email me if you want to learn more (email in profile). I'm not an expert on consumer data, but used to source business-related data feeds in a past life and got a glimpse of the consumer data side of things.

Zero PII. On occasion there are loyalty programs and IDs but there is no link back to the consumer. Companies like IRI and Nielsen recruit their own panelists and build models of the consumer that they then project into census style data.

Yeah, i don't know about that. There is a data feed I know of that contains the last 4 credit card digits...

Think of it in terms of checkout data: a transaction ID and itemlist with UPCs. Customer data is a whole 'nother enchilada.