When Do Apps Fire the ATT Prompt in Onboarding?

Across 12 tracked flows with an ATT prompt in Lazyweb Research's corpus, the ATT ask sits at an average step of 5.9 (median 2) inside longer flows that average 9.6 steps — about 70% of the way through [1]. ATT lands later and in longer flows than notification or location asks. The pattern: ask for tracking after the user is committed, not on the first screen.

ATT prompts sit at an average step of 5.9 in flows averaging 9.6 steps — later than notification (2.9) or location (2.8) asks — Lazyweb Research, July 2026.

Lazyweb Research · n=12 · Published 2026-07-07

onboardingmobileux-patternsmonetizationexperiments

The finding

Across 17 occurrences in 12 flows, ATT prompts land at an average step of 5.9 (median 2), average relative position 0.701, in flows averaging 9.6 steps [1]. The gap between the mean (5.9) and median (2) signals a split: some apps fire ATT early, but a tail of apps push it deep into long flows, dragging the average up.

ATT vs other permissions

PermissionAvg stepAvg flow lengthRelative position
Notification [2]2.95.60.643
Location [3]2.84.60.616
ATT [1]5.99.60.701

ATT sits later in absolute steps and in the longest flows [1][2][3]. But the counter-examples matter: Weight Watchers fired the native ATT dialog over the welcome/auth screen before signup — tracking asked pre-value [4].

How to apply it

Defer ATT until after the user has invested — completed signup, seen core value — so the tracking ask reads as 'improve what you already like' rather than a cold gate. The corpus's longer ATT flows support this. Asking ATT pre-value (as Weight Watchers does over the welcome screen) is an observed but riskier minority approach [4].

Caveats

This rests on 12 flows and 17 occurrences — small; lead with counts, not decimals [1]. The mean is skewed by a few very long flows, so the median (step 2) is the more robust central value. Positions come from ROW_NUMBER over captured screens, not guaranteed full user journeys.

The numbers

StatComputed from
ATT avg step 5.9, median 2, relative 0.701, flows average 9.6 steps; 17 occurrences across 12 flowsatt_position_in_flows: ROW_NUMBER-ranked flow_steps joined to ATT label set
Notification avg step 2.9, flows average 5.6 steps, relative 0.643notification_position_in_flows
Location avg step 2.8, flows average 4.6 steps, relative 0.616location_position_in_flows
Weight Watchers fired the native ATT dialog over the welcome/auth screen before signupqualitative: weight-watchers ATT over welcome screen
Methodology. Universe: 12 tracked flows (of 2,468) containing an ATT prompt, July 2026. Positions via ROW_NUMBER; small sample so counts and the median lead over the skewed mean.

Sources & citations

  1. [1] Lazyweb Research analysis of 12 flows (ATT positions, 2,468 tracked flows), July 2026. Small flow sample; ATT positions via ROW_NUMBER. Mean skewed by long flows, median step 2.
  2. [2] Lazyweb Research analysis of 26 flows (notification positions), July 2026. Comparison baseline.
  3. [3] Lazyweb Research analysis of 10 flows (location positions), July 2026. Comparison baseline; small sample.
  4. [4] Lazyweb Research qualitative review (807-app corpus), July 2026. Weight Watchers named as a pre-value ATT example.

Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.

Related questions

Explore the underlying screens, flows, and A/B tests inside Lazyweb. More research