100 installs as a test, then 1,700 for the top 5
A photo and video editor launched with almost no indexing. A hundred installs were bought as a test, and the answer set the size of the campaign.
A photo and video editor, a month old, in the Photo & Video category on the App Store. It was indexed for four keywords in its first market and collected about 75 organic installs a day. The conversion from impression to install was 6.8%, which is ordinary for the category, so there was nothing to fix there: the app was not losing people who saw it, it was not being seen.
Why the test came before the campaign
The usual advice is to start with low-volume keywords and work up. Instead the campaign opened with a hundred keyword installs spread across terms of every search volume, bought to answer one question: does this listing respond at all, and where.
It answered on the first day. Several keywords moved by close to 200 positions and landed in the top 50, and the high-volume terms, the ones that normally refuse to move on small volume, moved as well. That is what turned a test into a full campaign, and it is why the hundred installs were worth buying before the seventeen hundred.
What the position cost
Six days and roughly 1,700 installs took the app into the top 5 for photo editor and video editor, the two terms the whole category competes for. Before the installs started, its average position across the target keywords was 142.
Mid- and low-volume terms were much cheaper: the top 3 on 200 to 400 installs each.
The campaign stopped short of first place deliberately. Getting from the top 5 to number one on terms of that size costs more than the app could justify while its monetization was still being worked out, and by then organic installs had reached 250 a day on their own.
The numbers at the end of the run
Indexing in the first market went from 4 keywords to 100 after the metadata was rewritten. Daily organic installs went from 75 to 250, which is 3.3 times, over three weeks.
The cost figure needs its formula stated, because it is not a price per install we sell. The team divided what the incentive installs cost by the increase in organic installs, and got $0.50 per extra organic install. On the same arithmetic they expected about $0.30 once fewer installs were needed to hold the positions, which is a forecast made at the time, not a result anyone measured.
What this case does not show
Holding the positions was not free. High-volume terms needed repeat boosts, and the plan called for around 350 installs a day to keep them, so “organic growth” here means growth that is being maintained, not growth that was let go and stayed.
There is no revenue, retention, or subscription number in it. The app's monetization was still being built while the campaign ran, which is exactly why the campaign stopped where it did.
And the metadata did half of the work. Four indexed keywords became a hundred before a single install was bought. Installs move a position; they do not put an app into results it does not match.