I’m chairing the final paper session for today at the 5th Data Donation Symposium at the Weizenbaum-Institut in Berlin, which starts with Hekmat Alrouh; his interest is in the study of grocery shopping behaviours in Dutch supermarkets, and this will involve data donations of loyalty card information.
To test this idea, he began with a GDPR request to two supermarkets for his own shopping data; those data packages arrived two weeks later, but they looked very different from what he could see on his loyalty apps. He then contacted the higher-level executives in these corporations directly, by guessing their email addresses from email naming patterns, to flag this GDPR-noncompliance with them.
This produced a direct response from the Head of Legal at one of the supermarket chains within 24 hours. He replicated this immediately with the second supermarket, and also immediately received a full data package of his three-year shopping history. Organisational styles remained different, though – at one organisation, he is continuing to engage with their legal representative only, while at the other, he is conversing with the entire technical team.
There are still some open issues to be addressed, including personally identifiable data, location details, and other aspects; and details of the responsible parties on the university side. Corporate response patterns to this can be explained by organisational psychology and organisational structure; frontline workers have very limited agency, while higher-level executive staff will ask their teams to solve problems immediately.
So, being donor zero can be a very valuable strategy in developing the data donation path; escalating early and high can also be highly beneficial. Addressing commercial fears is also a useful approach here; and this also helps negotiate the route and not only the data in the data donation process.
To scale up from loyalty card information to household diets still requires some gaps to be addressed, of course: how often do households dine out of house; where does the rest of their food come from; what is the market share of the grocery chain; how complete are the shopping records; and how can placeholder rows make these omissions countable? The loyalty card alone is not representative of the household.
Data donation projects also need to address the question of what households gain from their participation: can they see whether they use shopping bargains effectively; can they identify their key shopping habits; can they assess how their patterns diverge from those of other households? Such insights might motivate participation. Auditing donor routes before they use them is critical, then; coverage of data donations also needs to be measured; and giving data back to donors motivates and rewards their contributions.
