The first message problem
Why outreach response rates collapsed, and what a message that still works looks like in 2026.
Reply rates to cold recruiter outreach in European tech have roughly halved over the last four years. Senior engineers in Amsterdam and Berlin report twenty to forty approaches a month. The rational response to twenty near-identical messages is to read none of them.
Most advice at this point tells you to personalise. That advice is not wrong, it is just too vague to act on, and it has produced a generation of messages that open with "I saw you worked at Adyen , impressive!" which is personalisation in the same way a mail merge is a letter.
The test
Before you send anything, ask: could this message have been sent to fifty other people with a name swap? If yes, it is not outreach, it is spam with a haircut.
The bar is not "mentions something about them". The bar is "contains information the recipient could not have predicted".
What that looks like
Weak:
Hi Sofia, I came across your profile and was impressed by your experience at Klarna. We have an exciting opportunity for a Senior Data Engineer at a fast-growing fintech.
Better:
Hi Sofia, you wrote a comment on the dbt Slack about ditching incremental models for a full-refresh pattern on smaller warehouses. We're doing the opposite and it is causing us pain. The role is data engineering at a 90-person insurer in Utrecht, 85-100k, hybrid two days. Happy to just talk about the modelling thing if the job isn't interesting.
The second one takes eleven minutes to write. That is the actual constraint, and it is why the first one keeps getting sent.
Say the money
Pay transparency rules are arriving across the EU regardless, and candidates already filter on it. Omitting the range signals either that it is bad or that you intend to anchor low. Both are read as a reason not to reply.
Give them an exit
The clause that raises reply rates most reliably in our experience is the one that says they do not have to be interested. "If the timing's wrong, tell me and I'll stop" converts non-answers into answers, and answers into a pipeline you can trust.
Measure the right thing
- Reply rate, not open rate.
- Positive reply rate, separately.
- Replies per hour of recruiter time , the number that decides whether personalisation is worth it.
Most teams track messages sent. That is the metric that produced the problem.