Update, 30 October 2026: after this went up, I read the requests the workflow had actually sent. Every one still had placeholders where my name, address and email should have been, so none of them could be acted on as a real request. The emails went out and the replies came back, but the workflow did not do its job. I am writing up what the organisations did with them.
Somewhere, a company you have never heard of holds your name, your address, probably your phone number, and a guess at your income and your interests. It bought them, or scraped them, and it sells access to them. That company is a data broker, and there are a lot of them.
Under UK GDPR you have real rights over what they hold. The trouble is that exercising those rights is a chore designed, whether on purpose or not, to be given up on. So I automated it.
The rights involved
Three articles do most of the work:
- Article 15, the right of access. You can ask an organisation to confirm whether it holds your personal data, and to give you a copy of it along with what it uses it for and where it got it.
- Article 17, the right to erasure. You can ask it to delete your data. The right is not absolute, but for a broker that holds your details purely to sell them on, the grounds for refusing are thin.
- Article 21, the right to object. You can object to your data being processed, and where the processing is for direct marketing, the objection is absolute: once you object, they must stop.
An organisation normally has one month to respond, which can be extended by two further months for complex requests, and it cannot usually charge you for doing so.
None of that is hard to understand. The hard part is doing it dozens of times.
Why doing it by hand fails
For one broker, a request is an email. For every broker that plausibly holds your data, it is a project. You need the right contact for each, which is often buried in a privacy policy and sometimes wrong. You need to write a request that cites the right articles and asks for the right things. You need to track which ones have acknowledged, which have asked you to verify your identity, which have gone quiet, and which deadlines have passed.
It is exactly the kind of work that gets started with enthusiasm, done for six brokers, and then abandoned in a folder. Repetitive, rule-based, deadline-driven, and entirely made of email: in other words, a textbook automation.
What the workflow does
It is an n8n workflow built around a list of brokers and their verified privacy contacts. For each one, it sends a request that covers access, erasure and objection together, in plain terms, citing the relevant articles.
A few principles shaped it:
- Verified contacts only. Every request goes to an address confirmed from the broker’s own privacy policy or published contact details. A request sent to a guessed address is not a request, it is a bounce, and a bounce starts no legal clock.
- One request, three rights. Asking for a copy, deletion and an end to processing in a single message is cleaner for both sides than three separate threads.
- Everything is a thread. Each request lives as its own email thread, so every acknowledgement, identity check and eventual answer lands in the same place, and the record of what was asked and when is the thread itself.
Did it work?
Not in the way that counts. The requests went out to verified contacts, and acknowledgements came back, but as the update at the top explains, the requests themselves still had placeholders in place of my details. By late August my inbox held 28 threads of this correspondence, filed under their own label.
Not every reply can be handled by a workflow, and it is worth knowing that before you build one. An organisation is allowed to ask you to verify your identity before it acts on a request. That is legitimate, since a broker handing your data to anyone who asked for it would be its own privacy problem, but it means some threads need a person to finish them.
The tool that is still looking for its market
Here is the honest part. The workflow runs, and I still have not decided what it is for.
There are two directions, and they are quite different businesses:
- A tool for individuals, or a free lead magnet: a way for someone to send these requests without spending an evening on it. Easy to explain, easy to want, hard to charge for, and heavy on support, because every reply that needs a human is a reply someone has to handle.
- A tool for businesses on the other side of the table: the companies that receive these requests and have to answer them within a month. Handling subject access requests is a real, recurring, legally deadlined cost for any organisation holding personal data, and the same workflow skills apply in reverse: log the request, start the clock, gather the data, respond, keep the record.
The second is the better business and the less exciting story. I have not picked yet, and I would rather say so than pretend the positioning is settled. Building the thing first and finding its market second is a mistake I have written about before, in the post about the ROAS dashboard. The difference here is that the build was small.
If you want to do this yourself
You do not need a workflow to exercise these rights. For any organisation you think holds your data:
- Find its privacy contact, usually in the privacy policy.
- Send one email asking, under Articles 15, 17 and 21 of the UK GDPR, for a copy of your data, its deletion, and an end to any processing including direct marketing.
- Keep the thread, note the date, and expect an answer within a month.
- If it does not come, the Information Commissioner’s Office is who you complain to.
If you are a business on the receiving end of these requests and the one-month clock is starting to hurt, I would like to hear how you handle them now.
This post describes rights under UK GDPR as I understand them. It is not legal advice.