Follow-Up Is the Whole Job

Ask a business owner why a particular deal did not happen and you will usually get a reason: the price, the timing, a competitor, a change of priorities. Ask them to walk through what actually happened, message by message, and a different answer often emerges. Somebody said they would send something. It did not get sent. Two weeks passed. By the time anyone noticed, the moment had gone.

This is not a failure of salesmanship. It is a failure of memory, and memory is a solvable problem in a way that persuasion is not.

Why silence looks like rejection from both sides

The customer's experience of being followed up with badly is not that you are disorganised. It is that you are not interested.

They asked a question. Nothing came back. The reasonable inference is that they are not important enough to answer, and the reasonable response is to stop thinking about you. Very few people chase a supplier who has gone quiet, because chasing feels like asking for a favour.

From your side, the same silence reads as a decision. They did not reply, so they must have chosen someone else. Both parties conclude the other has lost interest, and neither is right. This symmetry is what makes the failure so durable: nobody involved has any reason to investigate.

The administrative reality

Follow-up fails for reasons that are almost embarrassingly mundane.

The commitment was made in a conversation and written down nowhere. The person who made it was travelling. It was written down, but in a personal task list that nobody else can see, so when that person got busy there was no second line of defence. Or the follow-up was logged against a deal rather than a person, and when the deal was closed as lost, the reminder went with it — even though the person is still there, still working at the same company, and will have the same need again in a year.

The common thread is that the commitment lives somewhere private. A promise held in one person's head, inbox, or notebook is a promise with a single point of failure, and the failure mode is ordinary human busyness rather than negligence.

What a follow-up system actually has to do

It needs to answer three questions without anyone having to remember them.

Who is waiting on us? Not which deals are open — who is expecting something. These are different lists, and the second one is shorter and more urgent.

What did we say we would do? Specificity matters, because a vague reminder produces a vague message. "Follow up with Sarah" becomes "just checking in", which is a message that asks the recipient to do the work of remembering why you are writing. "Send Sarah the revised quote with the two-year option" produces a message worth reading.

When does this stop? A follow-up sequence with no end is a slow slide into being ignored. Deciding in advance how many attempts you will make, and what you will do when they are exhausted, is what separates persistence from nuisance.

The restraint that makes it work

There is a strong temptation to solve follow-up by capturing more: more fields, more logged activities, more automated touches. This tends to make things worse, because a system that demands a lot of input gets a little input, unevenly, and then cannot be trusted.

The discipline that keeps a follow-up record usable is the same one that keeps any record usable — not collecting data speculatively "in case it is useful later" [1]. A record with three fields that are always filled in beats one with twenty that are filled in sometimes, because the first can be relied on and the second has to be verified every time it is used.

Where automation earns its place, and where it does not

Automating the reminder is almost always right. The reminder is a memory problem, and machines are better at memory than people.

Automating the message is a different matter, and the line is worth drawing carefully. A sequence of generic emails sent to someone who had a specific conversation with you is not follow-up; it is a slower way of being forgotten. The recipient can tell, and what they learn is that the earlier conversation meant less than they thought.

If you do send automated sequences, two mechanical things matter more than most people realise.

The first is unsubscribing, and it should be genuinely one click. There is a whole specification for exactly this — a method "for signaling a one-click function for the List-Unsubscribe email header field" [2] — precisely because the alternative, making people hunt for a link and then confirm on a landing page, converts a mild disengagement into a spam complaint.

The second is that automated follow-up rides on the same domain reputation as everything else you send. Sending more, to people who want it less, is the fastest way to damage the channel you depend on for the conversations that are working. The reporting half of DMARC exists for this: it lets a sending organisation express policy and get back reports that a receiving organisation uses "to improve mail handling" [3]. If you are going to increase volume, read those reports rather than assume.

The follow-up nobody schedules: after you win

There is one gap so consistent it is almost a rule. Businesses build some kind of follow-up for prospects and almost none for customers.

The moment a deal closes, the person disappears from the pipeline, and with them goes every mechanism that was making sure they heard from someone. They are now in "delivery" or "support" or nothing at all — categories that are reactive by design. Nobody is scheduled to contact them; people respond when contacted.

This is where renewals quietly die, and where the second purchase that should have been easy never happens. The customer is not dissatisfied. They simply have no reason to think about you, because the only period in the relationship when you were reliably in touch was the period before they paid.

The fix is the same mechanism pointed the other way: a small number of deliberate, scheduled contacts after the sale, with a specific purpose each time — checking the thing they bought is working, asking what they would change, telling them about something genuinely relevant. Three a year is enough. Zero, which is the current number in most businesses, is not.

A modest test

Take the last twenty enquiries that did not become customers. For each, find the last message in the thread and note who sent it.

If more than a handful end with the customer, the problem is not your pricing, your product, or your market. It is that the conversation stopped on your side, and nobody noticed.

That is a much better problem to have, because unlike the others, it can be fixed this week.