How to tell whether a keyword campaign worked
What to check before, during, and after a keyword campaign: the baseline, the shape of the position over the run, and the traps that distort the result.
A keyword campaign has one job: move your app up for a keyword people search. So the test of whether it worked is the position for that keyword, read over the whole run and compared with what it did before. This post covers how to set that up and the mistakes that make a result look better or worse than it was.
Before the campaign: get a baseline
Start tracking the keyword before any installs arrive. A week of daily readings is enough to see how the position behaves on its own.
You need this because positions move without any campaign. In a month of daily snapshots of the US App Store top 10, we found that on an ordinary day about a quarter of the apps in a keyword's top 10 sat in a different place than the day before. Without a baseline you cannot tell a campaign's effect from that everyday movement.
Write down three things on the day you start:
- The position for each keyword in each country.
- Your usual daily installs from your own console.
- Any other change you are making at the same time: a new title, new screenshots, a release, an ad campaign.
The third one matters. If you change the title in the same week, you will not know which change moved the position.
During the campaign: read the shape
Look at the whole series, not two points. These are the shapes you will see most often:
| What the position does | What it usually means |
|---|---|
| Climbs while installs arrive | The campaign is moving the keyword |
| Climbs, then holds after the campaign stops | The app now earns enough installs at that position to keep it |
| Climbs, then slides back after the stop | The campaign moved it, and the app could not defend the place on its own |
| Stays where it was | The keyword is too competitive for this volume, or the app is not eligible for it |
| Sits at 1st to 3rd place the whole time | Nothing to climb. Here the result is that the position did not slip |
The last row trips people up. If the app was already near the top, a flat line means the campaign held the position.
After the campaign: watch for a few weeks
Keep the keyword on the chart after the last install. Holding a position earns organic installs, and those installs are what keep it. A position that slides back within days tells you the keyword was harder to defend than it looked.
Be careful with improvements that arrive late. In our casino campaign in Australia, three low-volume keywords entered the top 10 three weeks after the last install. That is too far out to claim, and we did not count it.
What you cannot measure this way
Installs, one by one. Campaign installs arrive mixed in with your usual traffic, so totals in App Store Connect or Google Play Console include both. Look at whether total installs rose during the run, and treat the position as the part you can check on your own.
Revenue. A position brings visibility. What people do after they install depends on the app.
Cause, with certainty. Rankings move for many reasons: competitors, store changes, seasons. A clean shape that lines up with the campaign dates is strong evidence. It is still not proof, so write it up as evidence.
Mistakes that distort the result
- Judging by one day. A jump of a few places overnight is within ordinary movement.
- Starting on the biggest keyword. In competitive categories, volume alone does not buy a position. In the casino case above, 1,260 installs left
pokiesbetween 200th and 216th place for the whole run. - Changing the listing mid-campaign. You lose the ability to tell what moved the position.
- Tracking only the keywords you paid for. Related keywords can move too. Keep them on the chart, and do not claim them unless the timing is clear.
- Stopping the chart when the campaign stops. The weeks after are where you learn whether the position holds.
How it works on Rankforge
- The baseline is free. Add the keywords and countries to the tracker before you buy anything. It records the position every day for each store, with no card needed.
- You set the run. You choose the keywords, the countries, the daily pace, and the dates, and you can stop a campaign at any point if the position is not responding.
- The chart keeps going after the stop. Tracking does not end with the campaign, so the weeks after the last install are on the same chart.
A short checklist
- Track every keyword at least a week before the start.
- Record the starting position, usual installs, and any other change.
- Read the position daily during the run, and judge it weekly.
- Keep tracking for at least three weeks after the last install.
- Report the shape, the dates, and what else changed at the same time.
Frequently asked questions
- What is the main sign that a keyword campaign worked?
- The position for that keyword in store search. It should climb while the installs arrive, then either hold or slide back after they stop. Install totals alone cannot show this, because campaign installs arrive mixed in with your usual traffic.
- How long before the campaign should I start tracking?
- Long enough to see how the position behaves on its own. A week of daily readings before the start gives you a baseline to compare against.
- Can I match campaign installs one by one in App Store Connect or Google Play Console?
- No. Installs from a campaign arrive alongside your usual traffic, so totals in your console mix the two. The position for the keyword is the part you can check independently.
- What if the position improves weeks after the campaign ended?
- Treat it with caution. Something else may have moved it. In one of our own cases, low-volume keywords entered the top 10 three weeks after the last install, and we did not count that as our result.