SYS/CASE
Growth & Performance · CRO · Analytics · 2022
Infirmière Reconversion — +150% conversion on the sign-up tunnel
By Mathieu Milon, AI engineer growth & CRO · Published
Skills assessment for nurses, more than 2,500 projects supported. The WordPress / Elementor site captures an organic audience; the Calendly booking sat too far below the fold.
+150% conversion
Context
Infirmière Reconversion supports caregivers who want to take stock, evolve or change career — a 10-week programme, coaches all former RNs, CPF-eligible.
The site and the booking journey
The site, on WordPress / Elementor, captures an organic audience via its blog, then points to a free appointment. The “Faire le point gratuitement” CTA in the hero and “RDV Gratuit” in the menu open a qualification pop-up, then the calendar.
Problem
Sign-up started too far below the fold: many visitors never saw Calendly. And when the form was too simple, it mainly attracted the curious — advisors ended up saturated with unqualified appointments.
Approach
Three changes, tested and measured over two months:
01 · CTA in the hero
“Faire le point gratuitement” CTA in the hero (70% of traffic) and “RDV Gratuit” in the menu, to capture intent on arrival.
02 · Calendly in a pop-up
Calendly in a pop-up — no more redirect to the booking page. A qualification filter first (“this call is for you if…”) then the calendar in place.
03 · Qualification filters
Qualification filters before the form: too simple = curious visitors, advisors saturated.
Filter before the call, not after: that is the heart of B2C lead generation, where volume only counts if it arrives qualified.
Tools
- Google Tag Manager
- Heatmaps
- Looker Studio
- Variant tests
Variant tests, Optimize-style, cover the hero and the form.
Where AI comes in
2022 engagement: heuristic and tests by hand.
What agents in the loop would change today
The tagging plan and the Looker funnel are the same gesture as Cassiopée — GTM, heatmaps, Looker to steer tests, not a decorative dashboard. Today, that tagging and those variants would be set and reviewed with agents in the loop.
Results
+150%
overall CVR in two months
Prospect quality
more mature candidates as iterations went on
| Measure | Before | After |
|---|---|---|
| Sign-up tunnel | 100 | 250 |
Base 100 = measure before intervention. +150% conversion in two months of tests; the hero CTA captures 70% of traffic.
Reporting dashboards
An HTML rebuild of the reporting stack shipped with the engagement: Google Tag Manager upstream, Looker Studio downstream. Each tab is a report; each report has its own pages.
Show the data
| S | Visits | Unique visitors |
|---|---|---|
| S09 | 4,620 | 3,540 |
| S10 | 4,980 | 3,810 |
| S11 | 5,240 | 4,010 |
| S12 | 5,410 | 4,140 |
| S13 | 5,690 | 4,350 |
| S14 | 5,320 | 4,070 |
| S15 | 5,480 | 4,190 |
| S16 | 5,760 | 4,410 |
| S17 | 5,760 | 4,420 |
Two visits in three come from a phone: the hero CTA had to fit the first screen of a handset, not of a 27-inch display.
Show the data
| Page | Visits | CTA click |
|---|---|---|
| /blog/reconversion-infirmiere-que-faire/ | 8,940 | 21.4% |
| / | 7,620 | 27.8% |
| /blog/bilan-de-competences-cpf/ | 6,180 | 19.1% |
| /notre-accompagnement/ | 4,350 | 24.6% |
| /blog/quitter-l-hopital/ | 3,890 | 16.2% |
| /temoignages/ | 2,470 | 14.8% |
| Band | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 09–11 | 42 | 48 | 45 | 51 | 38 | 22 |
| 11–13 | 56 | 61 | 58 | 63 | 49 | 31 |
| 13–15 | 49 | 52 | 55 | 50 | 44 | 28 |
| 15–17 | 63 | 68 | 66 | 71 | 58 | 35 |
| 17–19 | 84 | 88 | 91 | 86 | 79 | 52 |
| 19–21 | 92 | 95 | 97 | 93 | 88 | 61 |
The 5–9 pm bands saturate above 90%: nurses book after their shift. Two extra evening bands were opened in week 14.
One booking in two lands within 48 h. Past six days the no-show rate doubles — hence the SMS reminder.
Availability as the visitor sees it, inside the pop-up.
Show the data
| Sign-up funnel | n | % |
|---|---|---|
| Visits | 48,260 | 100.0% |
| CTA click | 9,180 | 19.0% |
| Pop-up seen | 7,240 | 78.9% |
| Slot picked | 2,060 | 28.5% |
| Booked | 1,128 | 54.8% |
The steepest step is still qualification filter → slot picked, at 28.5%. That is where the next test belongs.
| Funnel | Before | After |
|---|---|---|
| CTA click | 4.1% | 19.0% |
| Pop-up seen | 61.0% | 78.9% |
| Slot picked | 21.4% | 28.5% |
| Booked | 51.8% | 54.8% |
| Overall CVR | 0.93% | 2.34% |
Before = January–February, CTA below the fold and a dedicated booking page.
Show the data
| Where the +150% comes from | pt |
|---|---|
| CVR Jan–Feb | 0.93% |
| CTA in the hero | +0.72 |
| Calendly as a pop-up | +0.51 |
| Qualification filter | −0.24 |
| Micro-copy & SMS reminder | +0.42 |
| CVR Mar–Apr | 2.34% |
The qualification filter costs 0.24 points of CVR — on purpose: it hands advisors back the hours the browsers were taking.
Mock-up — demo data. The figures are fictional but internally consistent, anchored on the one public result of the engagement: +150% conversion in two months.
Takeaway
- 1
Raise the action: the best form does not convert if it is below the fold — the CTA has to live in the hero.
- 2
Remove the redirect: opening Calendly in a pop-up deletes a step and an opportunity to drop off.
- 3
Qualifying is converting better: a few filters protect advisors' time and raise appointment quality.
Visuals
Hero: CTA in the first screen, RDV Gratuit in the menu Booking pop-up: qualification filter, then calendar
Related expertise: CRO, Analytics, AI engineering.
A similar project?
Let's talk about your next engagement
A funnel, a product or an extract to analyse: book a 30-minute call, written scoping within 48h.