A huge challenge with data systems right now is “OK, great. So I have to do all of these things to comply with the GDPR and other privacy laws that are out there now (or will be soon, surely), but I have a conflicting obligation to provide useful information to my company, what do I do?”
Depending on your company, and your goals, there are many options on the table. One of the things we’ve been seeing being done more and more is to anonymize the data when someone removes their account.
For example, let’s say you have customers with activity records (could be orders or some other sort of interaction with your company and it’s offerings). You don’t want to lose the history of those interactions, but at the same time, the customer has asked to remove their information. It’s critical to think through what that means and what the customer’s goal is, then how to get there from here.
Too many people have been just out and out removing the account. I say “too many” because as soon as you outright delete accounts, you start to impair your ability to deliver useful reports and information to your company (or your clients).
Instead, since the customer’s real goal is not deletion, but the removal of their personally identifiable information consider other things that can be done to preserve the learning from their interactions, while still honoring their request and fulfilling your obligations.
– You could remove all personal information, replace it (name, address, phone, etc. – all that stuff) with “Private” or “Redacted” or whatever. By doing this, you update the master record, removing their information, but retaining the overall structure of your systems. Your reporting still works, you’re learning options still work, and the customer gets their information removed from the system. It’s a win-win.
– You can summarize the information in a data warehouse (the order information, other elements you legitimately need like a zip code, area code, etc.) and THEN remove the customer record. You still get all of the future learning options from your warehouse, but you’re of course honoring the removal request.
I’ve talked to so many people that are panicked about the whole data removal piece – we’ve been able to come up with some very straightforward options that honor what all sides want in these situations. Reporting on sales trends doesn’t care about individual names and such – and most of the corporate learning is more about that aggregate information. This is important to keep in mind as you set up your systems.
It’s thing brought up here a few times – “Seek first to understand….” except, in this case, you’ll need to understand the customer goals, the legal requirements AND your companies use and goals for the data. Then you can likely craft a rock-solid plan that doesn’t burn the proverbial house down to address the issue.