Some ask me whether they should turn 1 day view off on Meta. I keep it on when setting up a campaign, and it has nothing to do with making the numbers look prettier.
What 14 weeks of data shows
Across 14 weeks of one subscription app account, the chart groups every campaign by its attribution setting and sets Meta’s install count against the MMP’s.
The dashed marker is 1.0x, where Meta and the MMP agree on the install count.
Campaigns on click only came in at or slightly below the MMP. Turn 1 day view on and Meta runs 20% ahead, because the MMP’s view window was shorter than Meta’s. These are different campaigns grouped by setting, not one campaign with the setting flipped, so take the 20% as the size of the gap, not an exact price of the setting.
| Who counts the install | Counts a click for | Counts a view for | Meta installs ÷ MMP installs |
|---|---|---|---|
| The MMP, on every campaign | 7 days | 6 hours | 1.0x, the baseline |
| Meta, 7 day click or 1 day view | 7 days | 1 day | 1.20x |
| Meta, 7 day click | 7 days | Not counted | 0.97x |
| Meta, 1 day click | 1 day | Not counted | 0.91x |
The only Meta setting that runs ahead of the MMP is the one that counts views for longer than the MMP does.
What 1 day view buys me
The algorithm learns from more events. Cutting view attribution cuts signal, and on an account with limited volume, signal is the constraint.
The behavior it catches is also real. Someone downloads the app and never opens it. Someone installs, hesitates, then sees an ad that answers the thing they were unsure about, and opens. Neither leaves a click you can trace, and both can be the ad doing work. How often they are is a question for a lift test. On iOS, the same thing happens when people see the ad and search the app name later, and two reports show how much of your iOS organic is paid Meta traffic.
To close the reporting gap, increase the view attribution window in the MMP’s Meta configuration instead of cutting view attribution on Meta’s side. Matching the windows is also what makes Meta against the MMP a useful budget comparison.
| What changes | Turn 1 day view off on Meta | Raise the MMP’s view window for Meta |
|---|---|---|
| Reporting gap | Meta should land near the MMP, like the click only campaigns at 0.97x and 0.91x. | Closes it by matching the two windows. |
| Signal for the algorithm | Cuts it. | Keeps it. The algorithm still learns from the view events. |
| Installs with no click to trace | Meta no longer credits the ad for them. | Meta keeps crediting the ad for them. |
Both options shrink the reporting gap. Only the MMP change keeps the signal, so that is the one I make.
7 day click plus 1 day view is still my default.
A different gap is a Meta web to app campaign that shows clicks while AppsFlyer shows no installs. That one has its own causes, and I cover the checks in why Meta web to app shows clicks but no installs in AppsFlyer.