Instagram versus Facebook
Both built by Meta, both open links in their own browser, and one of them is dramatically worse at letting you leave.
9 August 2026 ยท PAD team
The measured gap
On 21,000 taps routed between 3 and 9 August 2026, after routing:
| App | iOS taps | Stuck permanently | Rate |
|---|---|---|---|
| 11,073 | 5 | 0.05% | |
| 7,003 | 121 | 1.7% |
Roughly forty times more failures on Facebook, with the same code doing the routing on both.
Why
Instagram ships an undocumented scheme, instagram://extbrowser/?url=, that hands a URL to the system browser. Facebook has no working equivalent, so there is one fewer route out and everything depends on the store scheme being accepted directly.
Since iOS 26 it got worse: some Facebook builds stopped marking the user agent, so detection written earlier sends that traffic down the wrong branch entirely.
What it means for budget allocation
If you split spend between the two placements and compare results, part of the difference is not audience quality. It is whether the tap could arrive at all.
Correct for it before concluding that Facebook audiences convert worse for your app.
Practical response
- On Facebook, show the manual button faster, around 1.4 seconds
- Treat unmarked iOS traffic as possibly in-app rather than as Safari
- Measure the fallback rate per placement rather than in aggregate
Per-placement arrival rates, so the comparison stops being a guess.