Personalized Paths, Better Checkout
For SeDomicilier (Tech), I redesigned a B2B domiciliation flow to be clearer and shorter. Goal: cut abandonment, explain the mail offer, and keep pricing 100% consistent through to payment.
Context
CHECKOUT REDESIGN (HIGH DROP‑OFF, CONFUSION ON MAIL/PRICING)
Scope
FIELD & REMOTE RESEARCH, UX/UI, PROTOTYPING, TESTING, HANDOFF
Results
-15% BOUNCE ON KEY PAGES
+9% SALES OF ADD‑ON SERVICES
Context
Our subscription journey is outdated, years of tech and design debt have piled up. Customer feedback points to friction and confusion, making the experience feel slow and unreliable.
Current state of the funnel
estimate
4-5 steps
Customization
8 steps
Payment
1 step
Informations
10 steps
Account creation
1 step
KYC
> 5
Hypotheses :
The user journey might be too long
Some steps may be unclear
The flow lacks intuitiveness
Sales shadowing workshop
Goal: listen to prospect–rep conversations to surface the most frequent questions tied to the purchase flow.
Findings:
Customers did not understand the mail-handling offer, couldn’t tell plans apart, and didn’t know which subscription was best for their needs
Users saw an early price estimate that became significantly higher at payment, causing loss of trust and confusion about where the extra costs came from.
Problem & hypotheses
User tests
Strong confusion on the mail offer (6/6 in tests didn’t distinguish options; 4/6 didn’t understand digitization).
Final price ≠ estimate: surprise at payment recap (6/6).
Flow long/redundant (5/6); inadequate qualification (city choice not required).
Hypotheses
Shorter, personalized flow ↓ cognitive load → ↑ completion.
Step‑by‑step price transparency + clear recap → ↓ drop at payment.
Contextual upsells (aligned with the project) → ↑ AOV without annoyance.
Problem & hypotheses
User tests
Strong confusion on the mail offer (6/6 in tests didn’t distinguish options; 4/6 didn’t understand digitization).
Final price ≠ estimate: surprise at payment recap (6/6).
Flow long/redundant (5/6); inadequate qualification (city choice not required).
Hypotheses
Shorter, personalized flow ↓ cognitive load → ↑ completion.
Step‑by‑step price transparency + clear recap → ↓ drop at payment.
Contextual upsells (aligned with the project) → ↑ AOV without annoyance.
Problem & hypotheses
User tests
Strong confusion on the mail offer (6/6 in tests didn’t distinguish options; 4/6 didn’t understand digitization).
Final price ≠ estimate: surprise at payment recap (6/6).
Flow long/redundant (5/6); inadequate qualification (city choice not required).
Hypotheses
Shorter, personalized flow ↓ cognitive load → ↑ completion.
Step‑by‑step price transparency + clear recap → ↓ drop at payment.
Contextual upsells (aligned with the project) → ↑ AOV without annoyance.
Explorations → Decisions
A. Entry to flow
Tested: (A1) single CTA vs. (A2) 3 project‑oriented CTAs.
Decision: A2 (Domiciliation / Domiciliation + Creation / Domiciliation + Transfer) → immediate personalization.
Trade‑off: more choice upfront but fewer useless steps later.
B. Status qualification
Screens for Transfer vs. Creation and BIC/BNC filter for micro‑businesses.
Decision: filter ineligible cases early; adapt upsells and content.
Trade‑off: richer business logic, but friction avoided later.
C. Mail offer
Variants: (C1) long text vs. (C2) comparison table + use examples.
Decision: C2 with crisp bullets (digitization, forwarding, SLA).
Trade‑off: consumes screen space, but comprehension ↑.
D. Upsells
From generic “push” to contextual recommendations (rules by profile).
Decision: conditional display; cap 1–2 proposals/screen.
Trade‑off: potential revenue vs. perceived intrusion.
E. Price & recap
Persistent basket recap with live total; frequency choice; fee breakdown.
Decision: end‑to‑end transparency.
Trade‑off: pricing complexity technically, but stronger trust.
Final solution
Core flow: 1) Choose project (3 CTAs) → 2) Status (Creation/Transfer, BIC/BNC) → 3) Mail offer (clear comparison) → 4) Contextual upsells → 5) Persistent recap → 6) Transparent payment.
Empty states: simple explanations + examples; redirect to help if ineligible.
Errors: precise messages, focus fields, keep user inputs.
Edge cases: ineligible micro‑businesses, payment frequency changes late in flow, removing an added upsell.
Lessons & next steps
Lessons: price transparency reduces drop‑off; targeted upsells lift AOV without irritation; the mail offer needs pedagogy.
Next: step‑level instrumentation (events), A/B entry (1 CTA vs. 3), mobile optimization, in‑flow mail FAQ.
Available