Which Apps Run The Most Signup And Onboarding Experiments?
- By Ali Abouelatta
- n = 4,814
- Published
- Updated
- 2 min read
- signup · onboarding · experiments · mobile · ux-patterns · saas
Short answer
Across 4,814 detected experiments spanning 276 companies (July 2026), signup and account-login screens account for roughly 376 labeled experiments — sign up (127), account login (167), and account setup (82). source[1] At the company level, ChatGPT (at least 18 signup), Chime (17), Disney (17), and Capital One (14) lead the detected signup work. source[2] These are observed variations, not confirmed A/B tests.
Sign-up, account-login, and account-setup screens carry 376 labeled experiments across the corpus (July 2026); ChatGPT leads companies with at least 18 signup experiments.
Bar chart · Signup experiments (lower bound)
Signup experiments (lower bound) — The leaderboard
One block is about 0.6 Signup experiments (lower bound)
- ChatGPT
- at least 18 [2]
- Chime
- at least 17 [2]
- Disney
- at least 17 [2]
- Capital One
- at least 14 [2]
- Bumble
- at least 13 [2]
- Adidas
- at least 12 [2]
- Delta
- at least 12 [2]
- Babbel
- at least 11 [2]
- Replit
- at least 13 [2]
- Zoom
- at least 10 [2]
The finding
Across the corpus of 4,814 detected experiments (276 companies), the labeled signup-adjacent categories total roughly 376 experiments: sign up (127), account login (167), and account setup (82). source[1] At the company level, the deepest detected signup programs are ChatGPT (at least 18), Chime (17), Disney (17), and Capital One (14). source[2]
The leaderboard
Signup-heavy companies span AI, fintech, and streaming.
| Company | Signup experiments (lower bound) |
|---|---|
| ChatGPT | at least 18 source[2] |
| Chime | at least 17 source[2] |
| Disney | at least 17 source[2] |
| Capital One | at least 14 source[2] |
| Bumble | at least 13 source[2] |
| Adidas | at least 12 source[2] |
| Delta | at least 12 source[2] |
| Babbel | at least 11 source[2] |
| Replit | at least 13 source[2] |
| Zoom | at least 10 source[2] |
How to apply it
These are the apps to study when redesigning a signup or onboarding flow — they have the most detected iteration on those screens, so their current designs reflect the most experimentation. The recurring detected direction across them (Zoom, Replit, Wattpad, Brainly) is email-first entry, promoting links to buttons, and leading with value before credentials. Cite the per-company volume; individual diffs are covered on the company pages. All figures are detected variations. source[1]
Caveats
All figures are observed variations with LLM-inferred rationale, not company-confirmed A/B tests — no lift is measured. source[1] Surface splits are lower bounds because screen category is unlabeled on 1,425 experiments. source[3]
The numbers
| # | Stat | Computed from |
|---|---|---|
| [1] | 376 labeled signup-adjacent experiments (sign up 127, account login 167, account setup 82) within 4,814 | top_experiment_screen_categories (sign_up 127, account_login 167, account setup 82) |
| [2] | Company signup leaders: ChatGPT 18, Chime 17, Disney 17, Capital One 14, Bumble 13, Replit 13, Adidas 12, Delta 12, Babbel 11, Zoom 10 | company_total signup splits: chatgpt 18, chime 17, disney 17, capital-one 14, bumble 13, replit 13, adidas 12, delta 12, babbel 11, zoom 10 |
| [3] | 1,425 of 4,814 experiments have no screen category | screen_category_null_on_experiments (1425/4814) |
Methodology
Universe: 4,814 detected experiments across 276 companies, July 2026. Extraction: screen-category and per-company splits (COUNT DISTINCT experiment_id) on observed diffs. Caveat: detected variations only; splits are lower bounds.
Sources
- [1]
Lazyweb Research analysis of 4,814 detected experiments (276 companies, ~800-app mobile corpus), July 2026
Labeled screen-category counts for sign up / account login / account setup.
- [2]
Lazyweb Research analysis of 4,814 detected experiments (276 companies), July 2026
Per-company signup/onboarding surface splits (COUNT DISTINCT experiment_id).
- [3]
Lazyweb Research analysis of 4,814 detected experiments (276 companies), July 2026
screen_category NULL on 1,425 experiments; splits are lower bounds.
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.
Explore the evidence
- Full 45-frame August 2026 capture
- Three captured screens · August 2026