In short: Four things move a visibility measurement without anybody touching your website: a model release, an index refresh, the session the question was asked in, and the question itself becoming more or less commercial.
Each has a signature you can recognise in the sheet, and recognising it is the difference between a defensible report and a story.
The test that separates all four from your own work is whether competitors moved with you. Yours moves you; theirs moves everyone.
The Test That Comes First
Before attributing any change to your work, look at what happened to the other sites named in the same answers.
If the whole cast changed, something on their side changed. If your presence moved while the same competitors stayed where they were, the change is more likely yours. This is why the sheet records every site named rather than only whether you appeared: without that column the test is impossible.
It is not proof, it is the first filter. A competitor could have done the same work in the same quarter. But a change that moves your presence and nobody else's is the only shape that is even a candidate for being caused by you, and most reported wins do not have it.
Four Things That Move It
-
Model release the largest, and datable
- Everything moves at once, across questions and often across markets
- The set of sites named can change substantially in a single week
- Datable against the platform's own announcement, which is why the sheet marks those rows
-
Index refresh partial and quieter
- Only questions that trigger a live search are affected
- Newer sources appear; the older cast mostly stays
- Usually visible in conventional search at the same time, which is a useful cross-check
-
Session history your own fault, and avoidable
- Answers drift toward what was asked earlier in the same window
- Fixed by a fresh session per run, which is why it is a rule rather than a preference
-
The question changing meaning slow, and easy to miss
- A question becomes more commercial, or a new product category absorbs it
- The cast changes gradually over quarters rather than at once
- The only remedy is reading the answers, not the numbers
Reading A Quarter That Moved
Work through it in this order and stop when something explains it.
Did a platform ship a model in the window? Check the marked rows first.
Did the whole cast change, or only your line? Did the answers get longer or start citing a different kind of source, which points at an index or a product change rather than at anything about you? And did anything you actually shipped land before the measurement, with enough time for a crawler to have returned?
That last one catches more false wins than the other three combined. Work published two weeks before a measurement has usually not been re-crawled, re-indexed and re-considered, so a rise in that window is almost certainly something else.
Reading A Quarter That Did Not Move
The more common case, and the one the report has to survive.
A flat presence line with a changed source list is not a flat quarter: the places the category is read from have moved, and that is the target list changing under you.
A flat line with an unchanged cast and a narrowing spread means the measurement got more stable, which makes the next comparison sharper even though this one shows nothing.
Say it before it happens. The lesson on goals put this in month one for a reason: association moves in quarters, so the first re-measurement after the work is likely to show very little. Stated in advance it is planning. Stated when the flat number arrives it sounds like an excuse, however true it is.
Key Takeaways
- Four things move a measurement without you: a model release, an index refresh, session history, and the question's meaning drifting.
- The first filter is whether competitors moved too. Yours moves you; theirs moves everyone.
- A model release is datable and announces itself. A question changing meaning is slow and gets misattributed to whoever just started.
- Work shipped two weeks before a measurement has usually not been re-crawled, so a rise in that window is probably not yours.
- A flat line with a changed source list is not a flat quarter; the target list moved under you.
Check yourself
Before you move on
Not scored, not recorded, and not part of the certificate. Both answers are settled by a sentence in this lesson, and the reasoning appears whichever option you pick.
- 01
Your presence rose this quarter. What is the first thing to check?
- 02
Work shipped two weeks before a measurement, and the number rose. What is the honest reading?
Was this course worth your time?