Follow-up messages that add something each time

A follow-up earns a reply when it gives the prospect something the previous message did not: a new piece of evidence, a sharper framing of their problem, a useful asset, or a lower-effort way to say yes. "Just checking in" and "bumping this to the top of your inbox" give them nothing, and they respond in kind. The fix is structural, not stylistic — you decide what each touch will add before you send the first one.

Why does "just checking in" fail?

Because it transfers all the work to the prospect. A check-in re-asks the same question with less energy than the original, and it signals that you have run out of things to say. The prospect has not forgotten you; they have deprioritised you, and a message that adds nothing gives them no reason to re-prioritise.

The pattern is nearly universal. Most firms stop at two follow-up touches while deals typically need five or more — and a large part of the reason is that nobody can think of a credible third message on the spot. That is a content problem, and content problems are solved in advance. The 90-Day Follow-Up Framework treats follow-up as a designed sequence rather than a series of improvisations, and this article covers the design of the messages themselves.

What counts as "adding something"?

Five categories cover almost every useful follow-up:

1. New evidence. A result, an example, a before-and-after relevant to their situation. Not a brochure — one specific proof point.
2. A new angle. The same service viewed through a different business problem. If the first message talked about pipeline volume, the second can talk about the cost of the manual work they are doing instead.
3. A useful asset. A checklist, a teardown, a benchmark figure they can use whether or not they ever buy. This is the touch that gets forwarded internally.
4. A lower-effort step. Replace "any thoughts?" with a specific, small ask — fifteen minutes, one question, or a self-serve option such as a calendar link. Booking links remove the scheduling back-and-forth that kills momentum at exactly this point.
5. A graceful close. The final touch names the silence, closes the loop politely, and leaves the door open. It routinely outperforms the touches before it, because it is the only one with a deadline built in.

The test for every draft: could the prospect get something from this message without replying? If yes, send it. If no, rewrite it.

How do you structure a five-touch sequence?

Plan it as a whole, then let the system run it. The mechanism looks like this:

• When the first message gets no reply after three working days, then touch two goes out carrying the evidence — one proof point, two sentences of context, same ask.
• When touch two gets no reply after four more days, then touch three changes the angle: a different problem, a different cost of inaction, and a smaller ask.
• When touch three is opened but not answered, then touch four sends the asset — value with no strings, plus the booking link as a quiet postscript.
• When touch four gets nothing after a week, then touch five closes the loop: "I'll assume the timing is wrong and stop here — if that changes, this link still works."
• When any touch draws a reply, then the sequence stops immediately and a human takes over. Automation that keeps selling after someone has answered does more damage than no follow-up at all.

Notice that each touch is shorter than the last. Persistence with escalating length reads as pressure; persistence with shrinking asks reads as professionalism.

Should follow-ups stay on one channel?

Usually not. A sequence that alternates a call attempt with the email touches typically outperforms either channel alone, because the channels catch different people at different moments — the trade-offs are covered in email, phone, or both: follow-up channel maths. The rule that matters here: whatever the channel, the touch still has to add something. A voicemail that says "just following up on my email" is the audio version of the bump.

Can AI write these for you?

It can draft them, and drafting is the smallest part of the job. What AI cannot do is know which proof point is true for your firm, which asset your prospects would actually use, or when to stop — those are decisions, and AI won't fill your pipeline; a system using AI will. Use a model to generate variants of a touch you have designed. Do not use it to invent value you have not got.

Write the five touches once, wire them into the sequence, and review the replies monthly. The firms that follow up five times are not more persistent people than you — they simply decided what to say while it was still easy.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks and what to build first.