The rules aeoo uses to decide whether a change helped, written down so you can check every verdict yourself.
DAY 0
A change goes live
BEFORE
The starting point
Its last 28 days on Google, and the whole site’s
AFTER
Check-ins at 14, 31 and 59 days
Each window starts after the change went live
Compared with the whole site
Your change, net of how the site moved
Enough data?
No
Not enough dataWaits for the
next check-in
Yes
The day a change goes live, aeoo saves the 28 days before it. Then it reads the same numbers again, each time against how your whole site moved.
One change, measured
New title and description for /services/it-support
Day 0 · live 3 Nov
1.8%
click rate in the 28 days before it went live
Day 14 · first read
2.3%
shown 380 times so far, too few to tell
Day 31 · second read
1.8% → 1.8%
while the whole site moved +4% in the same weeks
Day 59 · last read
3.0%
still helping two months on
One change, measured
New title and description for /services/it-support
1.8%
click rate in the 28 days before it went live
Snapshot
The 28 days before it went live, for the page and for the whole site.
2.3%
shown 380 times so far, too few to tell
First read
Usually too early. A page Google showed too few times says so.
1.8% → 1.8%
while the whole site moved +4% in the same weeks
Second read
Judged against how the whole site moved in the same weeks.
3.0%
still helping two months on
Last read
The final verdict. Measuring stops and the result stays on the change.
Each one carries its reason, on the change itself, on Results and in your client’s portal. Lime is kept for the one that means it helped.
Better than the whole site did over the same weeks.
Shows on Results, in the monthly report and in the bell.
Within what the whole site did over the same weeks.
Nothing to credit or blame.
Worse than the whole site did over the same weeks.
Revert puts the old version back.
The first read comes two weeks after it went live, then after one and two months.
Nothing is judged before that.
Google showed the page too few times for a fair comparison.
Checked again at each read.
A new title should earn more clicks. A new section should lift the search it was written for. So each kind of change is read on its own number.
Titles, descriptions and structured data
What we read
Click rate in Google
Compared with
The whole site’s click rate in the same weeks
A new section on a page
What we read
Position for the search it was written for
Compared with
How the whole site’s positions moved
A new page
What we read
Views in Google, or AI answers that cite it
Compared with
Nothing before it: it starts at zero
An answer for an AI question
What we read
AI answers that name you
Compared with
The same question’s answers before it went live
No change takes credit it didn’t earn, and none hides a result it didn’t like.
Changes on the same page within four weeks share the credit, and the result says so.
If a change hurt, “Revert” puts the old version back on Webflow and sends the change back for review.
A page Google rarely shows gets Not enough data instead of a guess, and is checked again at each read.
Taken from the code that writes each verdict. When a rule changes, this page and the changelog change with it.
Book a demo, or check what the assistants can read on your site today. Free, no sign-up.