First place held for 46 days on a crypto keyword
A crypto wallet in the UK took first place on Google Play for its main keyword and held it for 46 days, including nine days after the installs stopped.
This is the longest hold we have written up: 46 days in first place on the keyword the app was bought for. It is also the case where one of the three keywords moved for reasons we cannot account for, so both halves are here.
The campaign
A crypto wallet on Google Play in the UK. The run was 3,385 installs across roughly seven weeks, with gaps rather than a continuous daily rate, spread over three keywords.
The keyword that worked
crypto wallet is a mid-volume term in the category. Before the campaign the app moved around between 7th and 10th place on it, which is close enough to the top to be worth pushing and far enough down to be invisible in practice.
It reached first place about two weeks into the run and stayed there for 46 consecutive days. Nine of those days came after the last install was delivered.
That last part is the interesting one. Compare it with the antivirus campaign, where the position started drifting the moment the installs stopped and had shed three places within a week. Same mechanic, different outcome, and the difference is not the campaign. It is the keyword: a term that is not being actively contested does not push you off it the day you stop paying.
The keyword that moved late
crypto is a broader term. The app sat at 33 to 35 on it for the whole month before the campaign, without moving.
Partway through the run it jumped to 11, and a few days later to second place. After the installs stopped it came back down slowly, going 3, 3, 4, and 5 over the following days rather than falling off.
The keyword we cannot explain
wallet is the highest-volume term of the three. It sat between 190 and 205 for the entire campaign, so on the face of it the campaign did nothing for it.
Then, more than a week after the last install was delivered, it appeared at 4th or 5th place.
We have no mechanism that explains this. The campaigns had ended. It could be traffic bought through an account we were not looking at, it could be the app's own growth on the back of the other two keywords, or it could be the store re-sorting for reasons of its own. We looked, and we could not tell.
We are writing it down instead of leaving it out, because leaving it out would make the case look tidier than the data is. If we counted it, this case would claim three keywords moved. It claims two.
What this case does not show
No organic installs, no revenue, no retention. The platform reports the installs we delivered. What arrived on its own once the app sat in first place is exactly the number a buyer wants, and we do not have it.
No app name, no purchase cost. The keywords, the market, and the store are real.
No calendar dates. Durations and the order of events are exact; the dates are left out on purpose, because a keyword plus a market plus a date range identifies the app to anyone who cares to check.
No claim on wallet, per the section above.
What we take from it
A hold is a property of the keyword, not of the campaign. Before you buy a position, the question worth asking is who else is currently paying to hold it, because that is what decides whether you keep it after you stop.