The email nobody writes outperforms the one you agonise over.

These came out of my own email platform while writing this page. The flow figures cover the twelve months to that date; the campaign benchmark is the last ninety days. Click rate here means unique clicks against emails delivered — the ordinary definition, not a flattering one.

Almost nobody spends their time in that ratio. The campaigns get the meeting, the argument about the subject line and the Friday afternoon — five or six a week, every week. The welcome sequence gets set up once and then runs for years while everyone forgets it exists.

It would be flattering to say this is clever segmentation. It isn't, mostly. It's timing.

Somebody who joined your list this morning is thinking about you right now. They came looking, they made a decision to hand over an address, and the next email they get from you arrives while that's still true. Somebody on a Tuesday broadcast signed up eighteen months ago and is thinking about something else entirely.

That's the whole mechanism. The automated email isn't better written — mine certainly isn't. It just turns up at the only moment when attention is guaranteed, and it does that for every single person who ever joins, without anyone remembering to send it.

The old setup treated everyone who ever handed over an email address the same way. The rebuild split people on what they'd actually done — whether they'd bought, and what kind of thing they'd bought — and sent accordingly.

Measured email against email, the rebuilt sequence runs at 4.8% against 3.5% for what it replaced. That's about 1.4×.

That 3.5% is two old flows pooled, and I've pooled them because it's the least flattering defensible reading. Taken on its own, one of the two was running at 5.8% on a small sample — ahead of the rebuild. I don't think that one survives its sample size, but you should know it's in there rather than find it later.

Which is a real gain and a modest one, and I want to be careful not to oversell it, because there's a much more interesting number sitting next to it. The rebuild was worth 1.4×. Having a welcome sequence at all is worth 2.8×. If you're choosing where to put an afternoon, that ordering matters.

• The old numbers rest on small volumes. A few thousand emails between the sequences it replaced, against nearly twenty-five thousand for the current one. Small samples move around, so treat 1.4× as a direction rather than a precise figure.
• Segmentation isn't cleanly the cause. One of the sequences I replaced was already splitting on whether someone had bought. So the gain can't all be attributed to segmenting, and I'm not going to pretend otherwise.
• The platform records the numbers, not the reasoning. What I changed and why is my memory. The click rates are checkable; the story about them isn't.
• Click rate isn't revenue. It's the metric I can show you cleanly without publishing my sales figures. A click is a decent proxy for attention and a poor one for money.

Two questions worth answering honestly about your own business, both of which take about five minutes to check:

• Does anything at all go out automatically when someone new arrives? If the answer is no, that's the highest-return afternoon available to you, and it isn't close. You're currently getting nothing from the one moment when people are guaranteed to be paying attention.
• If something does go out, when did anyone last look at it? Automated sequences rot quietly. Mine referenced a product we'd stopped stocking for longer than I'd like to admit, and nothing in the system was ever going to tell me.

Neither of those needs new software and neither is really an automation project. They're the sort of thing that turns up on an audit's list near the top, because the cost of not doing them is invisible and the fix takes an afternoon.