LinkedIn · 2026-07-22 · 6 min
Personalisation Is Not Flattery
Complimenting someone's job title is not personalisation, it is a template with a name field. Real personalisation is about relevance, not praise, and the difference decides whether a message gets a reply.
"Loved your recent post" is not personalisation. Neither is "Congrats on the new role" or "Impressive background in the industry". These are the phrases that make people distrust the word personalisation altogether, because they are technically true of the specific person and completely interchangeable with anyone else in a similar position.
Real personalisation is not about making someone feel noticed. It is about proving that the reason you are writing actually applies to them, specifically, and would not apply in the same way to the next hundred people on a list.
Why flattery reads as fake even when it is true
The strange thing about a flattering opener is that it can be entirely accurate and still fail, because accuracy is not the test the reader applies. The test is: could this line have been written before you knew anything about me except my job title. If the answer is yes, the compliment does not build trust, it destroys it, because it reveals the process behind the message rather than hiding it.
"Congrats on the new role" could be sent to anyone who changed jobs in the last month. It says nothing about whether you understand what that role actually involves at their specific company, and most readers know that instinctively, even if they cannot articulate why the line felt hollow.
What real personalisation looks like instead
Real personalisation connects a specific, verifiable fact about the person or company to a specific reason you are writing. It does not need to praise anything. It needs to prove relevance.
- Weak: "Saw you're the Head of Engineering at [Company], impressive!"
- Strong: "Saw [Company] posted about migrating to a microservices architecture last quarter. We have worked with a couple of teams who hit the same observability gap around month three of that kind of migration."
The second version contains no compliment at all. It works because it demonstrates specific knowledge and ties it directly to a reason for reaching out, which is a much higher bar than knowing someone's job title, and it is the bar that decides whether a message gets treated as worth a reply.
The test: could this be sent to 500 people unchanged
The fastest way to check whether a line is real personalisation or flattery in disguise is to ask whether it survives being sent, unedited, to five hundred other people in the same target list. A compliment about someone's title, industry, or "impressive background" survives that test easily, which is exactly the problem. A specific fact about a product launch, a technical decision, a hiring pattern, or a piece of content they actually wrote does not survive it, because it would be factually wrong for almost everyone else.
If a line survives the five-hundred-person test, it is not personalisation. It is a mail merge with good manners, dressed up to look like research when none happened.
Where the relevant details actually come from
Genuine personalisation takes research time, which is the actual cost that separates it from a template with a name field. The useful sources are usually:
- Company-level signals: funding announcements, product launches, job postings that reveal a new initiative, technology changes visible on their engineering blog or job listings.
- Individual signals: something the person actually wrote or said publicly, a recent talk, a specific project mentioned in their own words rather than inferred from a title.
- Structural signals: the shape of their team, the tools they are known to use, a gap between their stated priorities and what a typical company at their stage usually has in place.
None of these require praise. They require someone to have actually looked, which is the entire point, and it is also why this kind of research does not scale down to zero cost no matter what tooling produces the first draft of the message.
Why this matters more as AI-generated messages increase
As more outbound gets produced by tools that can generate a plausible-sounding compliment about anyone's profile in half a second, flattery-as-personalisation is becoming easier to spot and more thoroughly ignored, because recipients have seen enough of it to recognise the pattern instantly. The messages that still land are the ones that could only have been written by someone who actually read the thing they are referencing, which is precisely what a template cannot fake convincingly at scale.
This is also why personalisation does not scale the way volume-based outreach wants it to. Doing it properly for a hundred named accounts a week takes real research time from a real person, which is the whole reason it still works, and it is a large part of what we describe on our how it works page when we explain why campaigns are priced per profile rather than per message sent.
Next: take your last outbound message and delete every sentence that would still be true if you sent it to a competitor of the recipient. Whatever survives is your actual personalisation.