Seven workstreams shipped or moved this week across acquisition, activation, conversion and retention. Five are live and metering data. Two are in build for next week.
Every experiment this week targets the same funnel question: how do we move signups to paid faster, and keep them once they convert.
$1 offer May campaign. Five broadcasts shipped, three to go. Site banner and in-app banner cover Memorial weekend.
Onboarding dashboard v3 prototypes built. Synthetic testing run with Nanz before any user touches them.
11 in-app variants drafted on the reverse trial. Post trial series v2 launched. Calendly reminders live end to end.
Referral relaunch sent to existing staff and admins. Early signals look strong.
Post subscription drip live. Day 7 review ask split three ways, Day 14 referral ask.
Enable the variants, ship E3 to E5 on the $1 push, run the synthetic walkthrough on each onboarding bet.
Strapi marketing banner live through Sun 31 May. In-app banner active Sun 25 to Sun 31. In-app modal fires Thu 28. These cover the Memorial Day dead zone passively.
E3 peer / social proof on Wed 27. E4 48 hour countdown on Fri 29. E5 ends today on Sun 31. SMS1 last 24 hours Sat 30. SMS2 final hours Sun 31.
First name personalisation in subject lifts opens roughly six points. Carrying that pattern into every remaining send.
n91 sent to 1,407 against an 8,078 preview. Suspected CIO frequency cap dedup against n88. Resolving before E3 schedules.
Campaign 50. Design parity with controls so the test surface is messaging, not visuals. Live as drafts in CIO, ready to enable.
| Day | Variant | What it tests |
|---|---|---|
| 0 | Holdout | 50% see nothing on signup. Measures whether the day zero modal is worth keeping at all. |
| 1 | Help framing | Replaces invented onboarding hour with the winning Sales Outreach voice. Same Calendly destination. |
| 2, 3, 4 | Claim offer CTA | Same body as control, CTA verb swaps from Upgrade to Claim offer. Triple replication, 3x sample power. |
| 3 | Finish setup | Branch fires after reminders page viewed. Variant treats the user as past explanation and asks for the last step. |
| 7 | Halfway earned | Replaces conditional Day 7 of 14 framing with positive "earning their place" feature list. |
| 11 | Softer benefit | Keeps $1k payment hook. Swaps p1 to a vague benefit line. No hard commitments to settle times. |
| 12, 13, 14 | Calm timeline, tomorrow, today | Counterweights the loss aversion late emails with timeline reassurance. Day 14 drops the snooze CTA. |
Each variant fires a tracked response event named RT-msg-N-v2-{slug}. Read against the matching control event after roughly 200 deliveries per arm. Holdout decision needs ~4 weeks of traffic.
Replaces Calendly's default reminders with brand-aligned CIO templates so demo no-shows have a fighting chance.
All subject lines use the meeting type and host name attributes for personalisation, with safe defaults if the data is missing.
Replaces Campaign 9 (four emails, 50% off, no signals). Three A/B tests, one engagement T/F branch, monthly rotating offer code resolved via Liquid snippet so the journey runs continuously across months.
"$1 a month, try Carepatron again." A/B on subject: benefit led vs trial recap led.
T/F on appointments_count > 0. Engaged variant leads with their strongest metric. Disengaged variant pivots to Calendly.
Peer inspiration. A/B on body: 3 short vs 5 longer examples.
Feature focus. No test, single template.
A/B on channel: email only vs email plus matched SMS 4 hours later.
Calendly link, no code. Personal urgency only, no calendar urgency.
Conversion event: subscription_started, 28 day window. Same definition as v1 for clean comparison. No signals yet, journey just started.
Strip away everything. Create client, run appointment, get paid. Setup tasks defer to a drawer. Lowest eng cost.
One question at the door splits the user. Switcher path leads with CSV import. Greenfield leads with create first client.
Full screen chat where Ava asks questions and renders inline forms. No competitor does this. Highest eng cost.
Lands on a populated demo with 3 fake clients. "Make it yours" claims the workspace. Highest wow, hardest build.
Synthetic testing with Nanz in progress. Each prototype walked through under 60 seconds. Side by side comparison drives the bet for the build.
Reactivating the dormant referral lever with a clearer reward, cleaner message and a matching in-app modal.
Watching: referral page visits, link copy events, and downstream paid conversions over the next two weeks before deciding on cadence.
Triggers on entering the Paying customers segment. Captures reviews while the buying decision is fresh, then asks for a referral once the workspace is in active use.
Entry on first qualification into the Paying customers segment. Single shot per customer.
Random cohort branch by profile. 34% G2, 33% Capterra, 33% Trustpilot. Subject line "How's Carepatron working out so far?" Each variant routes to its own review surface.
Subject "Know a practitioner who'd like Carepatron?" Picks up where the review ask left off. Mirrors the $300 reward used in the standalone relaunch.
Customer continues in the steady state lifecycle. Re-enters the loop only on a new qualifying subscription.
Why the split: spreads review weight across the three platforms that matter for B2B SaaS healthcare buyers. Conditional send rules guard against asking customers who've already reviewed.
Memorial Day dead zone covered passively. No email into the long weekend. Banner and in-app carry the offer through.