Expect to be new again, mostly

Returning after months away is not the same as starting, and it is much closer to starting than most returners expect. The reviews survive; the visibility does not, and visibility is what was producing the bookings.

The useful frame is that you are a new listing with an old reputation attached, which is a genuinely better position than an empty profile and a considerably worse one than the profile you left.

What carries over and what does not

The asymmetry is the point. The things that took months to build and were slow to acquire - reviews, a rubric, a worked example - are exactly the things that survive. The things that were fast to acquire, mostly ranking and momentum, are the things that went.

That is a reasonable trade, and it means the recovery curve is steeper than the original climb. It does not mean it is quick.

Re-measure before you reprice

Two numbers moved while you were gone, and you cannot assume either direction.

Your minutes per job. Speed in this trade is a practised thing, and it decays like any practised thing. Do one complete job on practice material and time it honestly, exactly as you did the first time - the practice run exists for this, and it is more useful on the way back than it was on the way in, because you now have an old number to compare against. The old number is a check on optimism, not a forecast: Buehler, Griffin and Ross (1994) found people's time predictions ran consistently optimistic, and that the bias disappeared in one study when participants were told to connect the prediction to relevant past experience. Expect the first few jobs back to take somewhere between a quarter and half again as long as they did, and expect that gap to close within a fortnight rather than a month.

The market band. Prices in a category drift, and they do not drift uniformly. Crowding pushes down whatever is easiest to supply while leaving scarce formats alone, so a six-month absence can leave your written tier undercut and your recorded tier underpriced at the same time. Reading the market band without guessing is the method, and three profiles is still not a market.

Only after both numbers exist should you touch the card. Repricing on the way back from memory is how a returner spends a month at a rate that was correct last year.

The order to rebuild in

Visibility first, price second. This is the reverse of what returning earners usually do, because the price is the thing you have opinions about and the visibility is the thing you cannot see.

Take a few jobs at or just below your last rate to restart the activity signal that ranking runs on. Marketplace sorting rewards behaviour it can measure - recent completions, response time, delivery against stated turnaround - rather than anything you write, which is why how ranking works inside a marketplace matters more in the first fortnight back than the copy does.

Then raise, once the queue exists to justify it rather than before.

Update the profile before any of it. A listing that visibly has not been touched reads as an abandoned one, and buyers check last-active dates more carefully than they read the description. A dated line saying you are taking work again does more than a rewrite.

How long it actually takes

There is no published figure for this and anyone quoting one has made it up, so take direction rather than magnitude.

What returners consistently report is that the first booking takes longer than the fifth, that the fifth arrives much sooner after the first than it did originally, and that the whole thing runs on the order of weeks rather than the months the cold start took. The reason is mechanical: the loop that was slow to start the first time is being restarted with reviews already in it, so each completed job compounds immediately instead of after ten.

The failure mode is repricing in the third quiet week and concluding something from it. A fortnight of silence is a fortnight.

What changed while you were away

Two things move fast enough to matter across a break.

Automated assessment gets cheaper and better on the generic end continuously, so whatever part of your offer a model could produce is worth less than when you left - the technical account of what automated scoring reliably does is the map of which tier of your card has quietly moved. The consumer-side tools packaging that scoring are also what your prospective buyers have already tried before they come looking for a person, which was less true a year ago.

Measurement conventions and the vocabulary buyers use in briefs also drift, and the method side is documented elsewhere rather than here. Reading a season of recent briefs before you write a word of your own listing is twenty minutes that saves a repositioning.

If you never fully closed

The best version of a return is the one prepared for on the way out: a dormant profile rather than a deleted one, exported records, and a payout method that still works. Winding down cleanly is what makes a return cheap, and the difference between the two exits shows up entirely at this end.

If you did close everything, you are starting again with the reviews as a memory rather than an asset, and the platform-side mechanics of relisting - verification, listing rules, what a new judge profile needs - are Rate Cock's own to document rather than a general subject.