When should you stop emailing a customer?

Sending more email is the cheapest growth lever a small business has, which is why it is the one most often pulled until it breaks.

The damage is not the unsubscribes. It is that the mailbox providers are watching, and the signals they weigh most heavily are the ones you create by sending to people who do not want it.

The arithmetic nobody does

Say a thousand contacts, and two hundred of them have not opened anything in a year.

Sending to all thousand feels free. It is not: the two hundred lower your engagement rate, which is an input to whether Gmail and the rest put the next campaign in the inbox at all. So the people who DO want your email are less likely to see it, because you also mailed the people who do not.

The unengaged segment is not neutral. It is a tax on the engaged one.

The cap, and why it lives in the engine

Five touches in seven days, enforced by a send-eligibility gate alongside marketing consent.

It is a ceiling, not a target. Most small businesses do better at one or two, and the right number is whatever your own unsubscribe and complaint rates support — which is a measurement, not an opinion.

The reason it sits in the engine rather than in a campaign setting is that a cap you can raise is a cap you will raise, in a launch week, with the best intentions. Every send obeys it regardless of which screen started it.

There is also a shared cost that makes this everybody's business: mail leaves on a sending pool whose reputation is shared with other tenants. One business sending hard enough to generate complaints degrades deliverability for businesses that did nothing. A cap in an engine is how that is prevented rather than requested.

Transactional is a different class, deliberately

Ten transactional message types exist, and they are not marketing. A receipt, a confirmation, a reset — these are not touches and are not capped like touches.

Keeping them in a separate class matters because the alternative is the slide every business is tempted by: adding an offer to a receipt, then calling a campaign a notification. The classes are separate in the engine so that sentence never has to be argued.

What the segmentation is for

Twelve lifecycle templates and segmentation across behaviour, industry, location, buying intent, engagement, lifecycle stage, predicted conversion and inactivity risk — and the honest use of all of it is sending to fewer people more relevantly, not finding more reasons to send.

Fourteen campaign metrics, every one derived from the records you supplied. No projected rates, no benchmark pulled from the air: if nothing was counted, nothing is shown.

When to stop

Three practical lines, and none of them needs a tool to decide:

• When somebody has not opened anything in a year, stop and say so — a last email asking whether they want to stay is better than a quiet year of being ignored.
• When your complaint rate rises after a change in frequency, the change was the cause.
• When you cannot name what the next email is for, it is for you rather than for them.

The question to ask

Not "how often should I send". Ask: what happens to my delivery when I mail the people who never open?

Any tool that cannot answer that is helping you send more and leaving you to discover the cost.