The day-one setup lift held for a second week: completions up about a third on every comparable step, and client creation is now significantly up. Card clicks stayed down 45%. Conversion dipped in release week, then returned to normal.
Zero-step users fell from 51% to 43%. The gain lands at 2 and 3 steps.
| Steps completed | Before | After |
|---|---|---|
| 0 steps | 51.2% | 42.6% |
| 1 step | 19.9% | 20.4% |
| 2 steps | 14.5% | 18.9% |
| 3 steps | 12.0% | 15.7% |
| 4 steps | 2.4% | 2.4% |
Locations up 8.6 points, services 6.4, import 5.4, payments 0.2.
| Step | Before | After | Change |
|---|---|---|---|
| Add locations | 36.0% | 44.6% | +8.6 |
| Add services | 28.6% | 35.0% | +6.4 |
| Import clients | 25.4% | 30.8% | +5.4 |
| Online payments | 4.4% | 4.6% | +0.2 |
Skipping is new: about 1 in 4 skips at least one step on day one, resolving the checklist instead of leaving it to nag.
| Step | Skipped |
|---|---|
| Import clients | 17.1% |
| Online payments | 15.5% |
| Add locations | 8.4% |
| Add services | 6.9% |
| Invite team | 5.7% |
Setup, payment intent and client creation won. Cards lost.
| Day-one behaviour | Before → after | Change | Read |
|---|---|---|---|
| Created a client | 25.0%→30.0% | ▲ +5.0 | significant, z 3.1 |
| Clicked connect Stripe | 4.7%→8.8% | ▲ +4.1 | significant, z 5.3 |
| Invited or added a teammate | 7.8%→9.5% | +1.7 | borderline, z 1.7 |
| Created a note | 18.0%→17.4% | −0.6 | within noise |
| Shared their booking link | 8.9%→8.0% | −0.9 | within noise |
| Clicked a recommended card | 23.5%→13.0% | ▼ −10.5 | significant, z 6.7 |
Teammate row counts each design's own invite actions: staff records before the 2 July invite revamp; invitations, direct adds and invite links after. The checklist's team step is not comparable across designs, so it is not shown.
Note creation recovered to baseline in week 2, easing the AI-notes discovery worry. The card decline itself has persisted for 2 weeks.
Top 2 cards take 4 in 10 clicks. The bottom 4 feature cards earn about 2 clicks a day each across all users: show them only when relevant.
| Card | Clicks | Users |
|---|---|---|
| AI Scribe (notetaker) | 137 | 89 |
| Import templates | 97 | 77 |
| Add branding | 84 | 57 |
| Booking link | 56 | 36 |
| Automate intake | 49 | 37 |
| Insurance billing | 35 | 28 |
| Automate reminders | 31 | 27 |
| Get paid faster | 29 | 21 |
| Start video call | 28 | 24 |
| Sync calendar | 25 | 22 |
| Customer story (tracked from 13 July) | 24 | 21 |
| ePrescribe | 7 | 3 |
Mobile is 21% of signups in both periods (n = 157 after release).
| Day one | Desktop | Mobile |
|---|---|---|
| Completed 1+ of the 4 steps | 49.9%→57.5%▲ +7.6 | 45.0%→58.6%▲ +13.6 |
| Created a client | 26.9%→32.2%▲ +5.3 | 18.9%→23.6%+4.7 |
| Clicked a recommended card | 22.2%→13.4%▼ −8.8 | 28.1%→11.5%▼ −16.6 |
| Skipped a step (new) | 22.7% | 28.0% |
Retention is activity-based: did the person use the product on day 3 or day 7 after their own signup.
| Active on | Before → after | Change | Read |
|---|---|---|---|
| Day 3, any product event | 10.6%→12.8% | +2.2 | trending up, not significant |
| Day 3, in-app navigation | 8.1%→10.4% | +2.3 | trending up, not significant |
| Day 7, any product event | 11.5%→12.2% | +0.7 | flat |
| Day 7, in-app navigation | 9.7%→9.8% | +0.1 | flat |
Day N is the 24-hour window N days after each person's own signup; incomplete windows excluded (after cohort n = 327 for day 7). Day 7 always lands on the same weekday as signup, which is why it runs slightly above day 3 in the baseline.
Week 1 was a real dip: 0.99% subscribed within 7 days against a 3.14% organic baseline. Week 2 snapped back to 2.43% at 72 hours, the top of the June band. That pattern reads as a launch-week anomaly, not a persistent effect.
| Signup week | Signups | Organic subs in 72 h | Rate |
|---|---|---|---|
| 1 June | 480 | 10 | 2.08% |
| 8 June | 517 | 2 | 0.39% |
| 15 June | 489 | 9 | 1.84% |
| 22 June | 583 | 11 | 1.89% |
| 29 June | 495 | 12 | 2.42% |
| 6 July (release) | 452 | 3 | 0.66% |
| 13 July | 452 | 11 | 2.43% |
Final check on 24 July, when week 2's 7-day and week 1's 14-day windows mature. If week 2's 7-day rate lands near baseline, the release-week dip closes as an anomaly.