Rank comparison
Before and after, on the same map
Put two scans of one keyword side by side and we work out what happened at every point between them. One matrix, one headline figure, one link you can send to the person paying for the work.
The renewal conversation, in one picture
"Rankings are up" is a sentence a client has heard from everyone who has ever billed them. A map that was red in March and is green in August is not a sentence, and it is not something you have to be persuasive about.
So we built the comparison as the results page rather than as a report feature. Choose the business, the keyword and two dates. What comes back is both grids, the movement between them, and a change matrix that says what happened at each of the checkpoints.
What a comparison looks like
Two example scans of one keyword, and the change matrix we compute from them.
What each cell can say
Four outcomes, and we show all four with the same weight.
Climbed, and by how much
A cell that improved carries the number of positions it gained. Colour shows direction and magnitude at once, and the glyph says the same thing again so the grid still reads in greyscale or to a colour-blind client.
Entered the results
The best cell there is: nowhere in the top twenty before, ranking now. On a wide grid a cluster of these along one edge is the clearest picture of expanded coverage you will get.
Dropped out
Ranked before, absent now. We show it as plainly as the gains. A comparison that only knows how to go up is one screenshot away from embarrassing the person who trusted it.
No change
Flat cells are marked flat rather than left blank, because an empty square is indistinguishable from a square we failed to measure.
Why we refuse to compare across keywords
It would be easy to let you diff any two scans, and the result would look completely convincing. Two different search terms have two different competitive fields, so the movement between them is not movement at all. It is a comparison of unlike things wearing the clothes of a trend.
The pickers therefore offer only scans of the term you have selected. You lose nothing real and you gain the ability to trust every number on the page.
Common questions
Can I compare two different keywords?
No, and that is deliberate. Comparing across terms produces confident, specific, wrong movement. The pickers only offer scans of the keyword you selected, so a comparison cannot accidentally be built from two different questions.
Do the two scans have to be consecutive?
No. Pick any two runs of the term. A March scan against an August one is the comparison most retainer conversations actually want.
Can I send a comparison to a client?
Every comparison is a URL, so the business, the keyword and both dates live in the address. Paste it into an email and your client opens exactly what you were looking at. You can also save a comparison so it is waiting for you next month rather than being rebuilt from the pickers each time.
What if the same day has two scans of the same term?
The pickers add a time to the labels, but only for the dates that actually collide, so a list of distinct days stays clean.
Is there a shareable image?
Yes. A comparison that genuinely improved can be turned into a square social card with the headline figure on it. A flat or negative comparison does not get offered one, because the card computes its claim from the data rather than assuming the answer.
Run the before scan now
A comparison needs two points and the first one is free. Scan the term today so that in ninety days you have something to put beside it.