Diagnosis
Email lifecycle audit in four passes: purchase and abandonment journey replayed on mobile and desktop, review of the quarter’s 2,058 abandonments, inventory of the email base (sources, consent, age) and review of existing sends. The 11% open rate does not indict the subject lines: it indicts the absence of sorting. An unsegmented base gets the same email, at the same time, whether it bought last week or visited in 2024.
| Step | Volume | Pass rate |
|---|---|---|
| Visits | 42,000 | — |
| Add to cart | 2,940 | 7.0% |
| Checkouts initiated | 1,764 | 60% of carts |
| Orders | 882 | 50% of checkouts |
| Cart abandonments | 2,058 | 70% |
| Overall conversion | — | 2.1% |
| Average order | €62 | €54,684 / quarter revenue |
| Emails captured without an order | 3,450 (12 months) | 0 contacted |
| Single monthly newsletter | — | 11% open · 0.3% click |
| # | Leak | Weight | Where it breaks |
|---|---|---|---|
| 1 | No cart recovery | 41% | 2,058 abandonments per quarter, zero email sent |
| 2 | Dormant leads never re-contacted | 26% | 3,450 emails captured in 12 months, no reactivation scenario |
| 3 | Single, unsegmented newsletter | 14% | One send a month to the whole base: 11% open rate |
| 4 | No buyer exclusion | 11% | The −10% code went out after the order: lost margin, customers trained to wait |
| 5 | No email revenue measurement | 8% | Impossible to say what email earns, so impossible to arbitrate |
- Two leaks invisible to the owner, and yet decisive: the cart was never recovered at the right moment (the first email went out 5 days after abandonment, when the need had passed), and the base held 3,450 consenting contacts nobody had reopened the file on.
Treatment
Four waves, 12 days, 4.5 person-days. Nothing to rebuild on the site: it all happens in scenarios and instrumentation.
01
Wave 1 — Instrument and exclude (days 1 to 3)
- Clean events: cart_abandoned, checkout_started, order_placed, email_captured, unsubscribe, with reliable timestamps.
- Automatic exclusions: any order cancels the recovery, any buyer leaves acquisition scenarios, any unsubscriber leaves everything.
- Double opt-in and consent traceability: source, date, form. The existing base is sorted by interaction recency.
02
Wave 2 — Cart recovery branch, 4 emails (days 4 to 8)
- Trigger: abandonment confirmed after 45 minutes of inactivity, sends at D+1, D+3, D+5 and D+8, in the 10am–12pm and 6pm–8pm windows, never on Sunday.
- Segmentation by amount: carts under €150 in the short branch (3 emails), €150 and above in the long branch (4 emails, with the 3-instalment payment argument).
- Contained discount: no discount before email 4, single code valid 72 hours, one use per customer. Given too early, the discount becomes the price.
03
Wave 3 — Dormant leads branch, 2 emails (days 9 to 11)
- Reactivation at D+30 and D+45: useful content in the first email, a binary choice in the second (stay or leave), −10% discount valid 7 days on exiting the branch.
- Quarterly reactivation for contacts with no open in 60 days: content, never a promotion.
04
Wave 4 — Measure (day 12)
- Revenue per email, credited to the last click over a 7-day window, de-duplicating orders touched by several emails.
- Systematic A/B on subject lines: 50% of the flow, 10,000 sends minimum before concluding.
- Review at D+30 and D+90: opens, clicks, recovery, unsubscribes, complaints per 1,000 emails.
Results
Results
Comparison: 90 days before / 90 days after go-live.
1.9%
of carts recovered by email, against 0 — demo scenario
+12%
orders with not one extra euro of ads — demo scenario
€2,190
monthly revenue attributed to email — demo scenario
| Metric | Before | After | Delta |
|---|---|---|---|
| Visits | 42,000 | 43,100 | +2.6% |
| Add to cart | 2,940 | 3,017 | +2.6% |
| Cart abandonments | 2,058 | 2,111 | +2.6% |
| Carts recovered by email | 0 | 40 (1.9%) | +40 |
| Dormant leads re-contacted | 0 | 3,450 | +3,450 |
| First orders from dormant leads | 0 | 66 (1.9%) | +66 |
| Orders | 882 | 988 | +12% |
| Average order | €62 | €62 | — |
| Quarterly revenue | €54,684 | €61,256 | +€6,572 |
| Cart sequence open rate | — | 46% | — |
| Cart sequence click rate | — | 11% | — |
| Unsubscribes per email | — | 0.32% | — |
| What email earns | not measured | €2,190/month | — |
Project yield: €3,900 excl. tax (demo pricing) for €2,190 of extra monthly revenue, paid back in under two months, without touching the site. The gain does not come from a better-written email: it comes from 2,058 carts and 3,450 contacts waiting for a scenario, not a rebuild.
How to read these numbers
These figures are simulated. Three rules we apply on a real client:
- Assumed low baseline. A single, well-written email recovers 0.6% of abandonments. The full sequence targets 1.9% because it stacks four moments and three arguments. On a shop that already has recovery in place, the realistic expectation is +0.4 to +0.8 points, not 1.9%.
- The long game. 1.9% on 3,450 dormant leads means 66 orders over a quarter, not in a week. Dormant-lead recovery is assessed over 90 days minimum.
- Public stop thresholds. Open rate below 30% on the first email, unsubscribes above 0.5% per email, any complaint: cadence or argument is revised before continuing.
The Nexperio method — email lifecycle in five steps
01
Journey replayed
Abandonment, return to the site, order, with a test account and a real address.
02
Abandonment reading
By step, by device and by cart value.
03
Base inventory
Sources, consent, age, open rate per segment. We do not contact what we cannot prove consented.
04
Branch scenarios
From the most profitable to the finest: cart, dormant leads, post-purchase, reactivation.
05
Proof before rollout
Then review at D+30 and D+90, with stop thresholds written in advance.
Illustrative demo case. Fictional brand “Atelier Sauge”, no real customer data, no real email was sent. Figures come from a demo scenario calibrated on public cart-recovery benchmarks.
Internal ticket : NEX-39 · All success stories → /cas-de-guerison