Small Numbers Still Talk

Ten visitors a day feels like nothing.

It isn't.

When the numbers are tiny there's a temptation to ignore them until they're bigger, on the theory that nothing meaningful can be learned from so little.

But small numbers are often more honest than large ones, because there's no averaging to hide behind.

Each visit is a person, and what a single person does can be looked at closely.

Ten people came.

Where from?

If eight came from the same place, that's where the effort goes next.

Did any of them stay more than a few seconds?

If not, the page isn't saying what the link promised.

Did anyone click through to the thing being recommended?

One click out of ten is a rate.

Zero out of ten is also information, and probably the more important kind.

The mistake is to wait for statistical significance before paying attention.

That's a luxury for later.

Early on, the job is to watch behaviour and ask why.

Why did that one email get a reply and the other four didn't?

What was different?

Not "was it a fluke" but "what would I do if it wasn't?"

Small numbers can be acted on faster, too.

A change to a page that's getting ten visits a day can be judged in a week.

The same change on a big site takes a month to read cleanly.

Early on, the pace of learning can be quick precisely because nothing is at scale yet.

That's an advantage that goes away.

So look at the ten.

Every one of them.

They're saying something.

The only question is whether anyone's listening.