How Elistit measures whether a change worked

Approve a change in the Plan and it does not vanish. It moves to Activity, sits at “Measuring impact…” for a while, and eventually comes back with a verdict. Here is exactly how that verdict is reached, because a tool that grades its own homework should show its working.

Two windows, fourteen days each

When a change is applied, the day it went live becomes the pivot. Elistit then compares:

• Before: the 14 days leading up to the change
• After: the 14 days following it

Views, sales and revenue are read for each window from your real Etsy listing stats. Nothing is modelled or estimated.

This is why a fresh change sits at “Measuring impact…”. There is no verdict until there is an after-window to compare, and the measurement runs as part of your shop’s daily sync rather than the moment you click.

The five verdicts

A few consequences worth knowing:

• A sale outranks everything. If the listing sold in the after-window, the verdict is “sale” regardless of what views did. Money is the outcome the shop runs on.
• The dead band is deliberate. Between 85% and 115% the answer is “no measurable change”, not a tiny percentage dressed up as a result. Etsy view counts wobble on their own; reporting a 4% move as a win would be noise sold as signal.
• “N views since” means no baseline existed. Usually a newly created listing, or one whose stats began after the change. It is not a stronger result than “views up” – it is a weaker one, because there is nothing to compare against.

What ”+$X earned” means

Your track record card says something like ”+$18.40 earned across the last 6 measured”. Read that as precisely as it is meant.

The figure is the difference between the two windows: revenue in the 14 days after, minus revenue in the 14 days before. That is a comparison, not a causal claim.

It does not prove the change caused the sale. A tag refresh followed by a sale might have been the tags doing their job, or a buyer who would have found the listing anyway, or a seasonal upswing that lifted your whole shop. Elistit cannot separate those, and does not pretend to.

What the number is good for is the pattern across many changes. One ”+$12 after a tag refresh” is an anecdote. Twenty of them, with the losses included, is a signal about whether this lever works for your shop. That is why the card always shows how many changes were measured next to the total – a single spiky sale should never read as proof.

Why the count of wins matters more than the total

“3 of 6 improved” is the honest headline. It includes the three that did not.

Nothing is filtered out for looking bad. Declines are measured, shown, and counted in the denominator, and a change that hurt gets an Undo button that says so plainly. A track record that only ever went up would not be a track record.

What it cannot tell you

• Whether a different change would have done better. There is no control group. Your listing is not run twice.
• Anything about a listing you edited yourself on Etsy. Only changes approved through the Plan are measured.
• Anything within 14 days. Ask sooner and the honest answer is “not yet”.

FAQ

Why does something still say “Measuring impact” after two weeks? The measurement runs on your shop’s daily sync. If the sync has not run since the after-window closed, the verdict lands on the next one.

Do declines get undone automatically? No. Nothing is reverted without you. A change measured as a decline is flagged and offered an undo, but the decision stays yours.

Does the before-window include other changes I made? It can. If you approved two changes on the same listing close together, their windows overlap and the verdicts are not independent. Spacing changes out on a single listing gives you a cleaner read.