PT

Home → Answers → The learning loop

How does the content improve on its own?

Every published post reports its own results back, and those results are written into your brand file. The next plan is built from that file, so each month’s content is shaped by what your audience actually responded to in the months before it. In Leanboat this runs weekly, without you configuring anything.

What most content operations do instead

A post ships, it does whatever it does, and nothing downstream changes. Next month’s calendar was written before last month’s results existed. The work repeats at a constant quality forever, which is why month twelve of a content calendar looks a lot like month one.

The four steps, in order

1 · Publish
The approved week goes out to your accounts.
2 · Measure
Each post's numbers are collected daily from the account it ran on: reach, saves, shares, comments, and where followers came from.
3 · Write back
Those results go into your brand file. Not a report you have to read, but a change to the document that drafts are generated from.
4 · Re-plan
The next month's plan is built from the updated file. Subjects that worked come back in new forms, and ones that didn't quietly stop appearing.

The step that makes it a loop rather than a report is the third one. Measurement that lands in a dashboard changes nothing unless somebody reads the dashboard and acts. Measurement that lands in the file the drafts come from changes the next draft whether anybody reads anything or not.

What we can and can’t claim about it

We can say what the loop does, because it’s built and running: results are collected daily and written back, and plans are generated from the updated file.

We can’t yet show you a measured improvement curve. The honest reasons: the engine’s oldest brands have been through a small number of cycles, and post-level metrics in our own system currently cover a recent window rather than the full life of every post. Any “posts improve by X%” we published today would be a number we’d built rather than measured.

So this page describes a mechanism and stops there. When there’s a curve worth showing, it will appear here with its sample size attached.

What you do

Nothing. There is no experiment to configure, no test to name, no report to read. You keep approving the weeks you like, and the rejections count as signal too.

Related questions

Questions

On the loop.

How does Leanboat's content improve over time?

Each published post's results are collected and written into your brand file, and the next plan is generated from that updated file. What your audience responded to shapes what gets drafted next, weekly, without you configuring anything.

What does it measure?

Per post, on the account it ran on: reach, saves, shares, comments, and where new followers came from.

Do I have to read reports or run tests?

No. The results change the next drafts directly rather than arriving as a dashboard you have to act on.

How much better does it get?

We don't publish a figure for that. The mechanism is built and running, but the engine's brands have been through too few cycles for a measured improvement curve to be honest. When there is one, it will appear with its sample size.

Month three isn't month one.

Every post teaches the next one, without you running the experiment.