Which Apps Run The Most Detected UI Experiments Overall?

Across 4,814 detected UI experiments spanning 276 companies (July 2026), DoorDash leads with 221 distinct experiments — 2.6x the second-place app, Wattpad (85). [1][2] The top of the ranking is dominated by high-frequency consumer apps, and 138 of 276 companies clear the 10-experiment bar. [3] These are observed before/after variations with inferred rationale, not company-confirmed A/B tests.

DoorDash leads the corpus with 221 detected experiments (July 2026), more than the next two apps combined; 138 of 276 companies have 10 or more.

Lazyweb Research · n=4814 · Published 2026-07-07

experimentssaasmobileux-patternsdesignlanding-page

The finding

Lazyweb Research detected 4,814 distinct experiments across 276 companies. [4] DoorDash is the runaway leader at 221 experiments, more than the next two apps (Wattpad 85, Replit 77) combined. [1][2] Of the 276 companies, 138 clear the 10-experiment eligibility bar — meaning roughly half the corpus is a serious enough experimenter to teardown. [3]

RankAppDetected experiments
1DoorDash221 [1]
2Wattpad85 [2]
3Replit77 [5]
3Brainly77 [6]
5Zoom74 [7]

How to read the ranking

Raw experiment volume reflects both testing intensity and how much surface area an app has. DoorDash's 221 is inflated by a large home/merchandising surface (at least 45 home experiments) [1], whereas paywall-only apps like Tinder (28, all paywall) rank lower despite concentrated monetization testing. Use the ranking to find who iterates broadly; use the per-company teardowns to see where.

How to apply it

If you are benchmarking your own program, this ranking sets the ceiling: the most-tracked consumer apps run dozens of detectable variations, and half the corpus exceeds 10. A team running fewer than 10 UI experiments over the tracked window is below the median of apps worth tracking. Pick a same-category leader from the teardowns and match its cadence surface-by-surface. [3]

Caveats

All 4,814 are observed variations with LLM-inferred rationale, not company-confirmed A/B tests — no lift is measured. [4] Counts use COUNT(DISTINCT experiment_id); the screenshots join would otherwise inflate totals. Volume rankings favor apps with more screens, not necessarily more disciplined experimentation. [4]

The numbers

StatComputed from
DoorDash 221 distinct experimentscompany_total:doordash (value 221)
Wattpad 85 distinct experimentscompany_total:wattpad (value 85)
138 of 276 companies have >=10 experimentscompanies_ge10_experiments (138/276)
4,814 experiments across 276 companiestotal_detected_experiments (4814); companies_with_experiments (276)
Replit 77 distinct experimentscompany_total:replit (value 77)
Brainly 77 distinct experimentscompany_total:brainly (value 77)
Zoom 74 distinct experimentscompany_total:zoom (value 74)
Methodology. Universe: 4,814 distinct detected UI experiments across 276 companies (COUNT(DISTINCT experiment_id)), July 2026. Method: before/after screenshot diffs with LLM-inferred rationale, ranked by per-company distinct-experiment count. Caveat: detected variations only, never confirmed A/B tests; volume favors apps with more screens.

Sources & citations

  1. [1] Lazyweb Research analysis of 221 detected experiments (DoorDash, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id); corpus-leading total.
  2. [2] Lazyweb Research analysis of 85 detected experiments (Wattpad, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id); second-highest total.
  3. [3] Lazyweb Research analysis of 4,814 detected experiments (276 companies, ~800-app mobile corpus), July 2026. 138 companies have >=10 distinct experiments (the page-eligibility bar).
  4. [4] Lazyweb Research analysis of 4,814 detected experiments (276 companies, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id) across all companies; detection window 2022-10-10 to 2026-07-04.
  5. [5] Lazyweb Research analysis of 77 detected experiments (Replit, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id).
  6. [6] Lazyweb Research analysis of 77 detected experiments (Brainly, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id).
  7. [7] Lazyweb Research analysis of 74 detected experiments (Zoom, ~800-app mobile corpus), July 2026. COUNT(DISTINCT experiment_id).

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