What a Buissonnière customer is really worth, what it costs to win one, and why the next euro in media should go abroad first.
€107k of Meta + Google in 2025 against €9.3M in sales: half the 2.4% an average company spends on paid media.
Belgian online customers spend 42% of that in the stores, where webshop ROAS never sees it.
After filtering out the extra demand of collection launches. France and the Netherlands are the most efficient.
Better customers than a year earlier. France, the Netherlands and Switzerland pay back their media within six months; Germany within the year.
Everything in this deck comes from Buissonnière's own systems and ad accounts. Revenue is net of returns and excluding VAT, unless stated otherwise.
| Source | What it contains | Period | Volume |
|---|---|---|---|
| Back office export "Chiffre d'affaires journalier par client et par magasin" | Every sale Buissonnière books: one line per ticket (a till receipt in a store, or an order on the webshop), with amount, point of sale and, where known, the customer | Feb 2024 – Aug 2026 | 274,093 tickets |
| Magento customer lifetime report | Every customer who ever ordered on the webshop: e-mail, country, date of first order, orders and amount paid | Mar 2011 – Sep 2026 | 65,397 customers |
| Meta & Google Ads | Ad spend per campaign and per day, assigned to a country through the campaign name | Jan 2024 – Sep 2026 | €231k spend in total 2024 €53k · 2025 €107k · 2026 €71k |
One line in the back office = one till receipt in a store or one order on the webshop. A return is a ticket with a negative amount. Illustration below: same structure, values made up.
| Date | Ticket no. | Excl. VAT | Incl. VAT | Point of sale | Customer ID | |
|---|---|---|---|---|---|---|
| 12/03/2026 | 00023-90001xxxxx | €61.16 | €74.00 | Webshop | 9xxxxx | m•••@gmail.com |
| 28/03/2026 | 00011-10004xxxxx | €90.08 | €109.00 | Store | 9xxxxx | m•••@gmail.com |
| 02/04/2026 | 00011-10004xxxxx | €-24.79 | €-30.00 | Store · return | 9xxxxx | m•••@gmail.com |
| 02/04/2026 | 00014-20003xxxxx | €45.45 | €55.00 | Store | — | — (anonymous) |
Pay in a store with a customer card, and the till records the customer ID and e-mail. The same e-mail is on the webshop orders and in Magento. That is how one person's online and store purchases come together.
Country, date of first webshop order. This defines a "new customer" and their cohort.
Matched as an encrypted code (hash): the analysis never handles names or readable addresses.
All their tickets afterwards: webshop and stores, returns included.
A strong omnichannel brand with a small media budget that follows the collections.
Average company: marketing = 7.7% of revenue, 30.6% of that is paid media → ≈ 2.4% of revenue. Buissonnière spends half.
Zalando: marketing = 9.3% of revenue.
Gartner CMO Spend Survey 2025 (402 CMOs, mostly large firms) · Zalando Q3 2025 results
Belgium took 42% of paid media in 2025 and 49% in 2026. Budgets peak at each collection launch and capsule drop, as they should in a seasonal business. Later in this deck: per extra euro, France and the Netherlands add about twice as many new customers as Belgium.
Lifetime value, measured across webshop and stores, by country and by cohort.
Buissonnière is a repeat brand: families buying for every season and every child. That is what makes acquisition worth paying for.
These are all-time figures from Magento: older customers have had more years to buy. From the next slide on, we follow only customers acquired since Feb 2024, measured over the same window and including their store purchases.
Cumulative net revenue per new online customer, months after the first order, webshop + stores.
Belgium almost triples its first order within 18 months. Abroad, value comes mostly from a larger first basket; repeat is lower but rising.
Webshop ROAS and Google Analytics see none of this.
Same person = same e-mail on the webshop order and on the store receipt (customer card). Anonymous store purchases are not counted, so the real store share is higher.
One in six "new" Belgian online customers had already bought in a store before their first webshop order (seen in the back office since Feb 2024). They are already loyal: going online is a step up in their relationship with the brand, not a new customer won by ads.
International customers start with a larger first order; France, Germany and Switzerland also return less. Belgian customers come back more often and also buy in the stores.
Longer term confirms it: median value of customers who have been with Buissonnière 2+ years (Dotdigital, spring-summer 2026 review) is €175 in Belgium, €158 France, €238 Netherlands, €290 Germany, €403 Switzerland.
Six-month value per new customer: the same season one year apart (September to February), so Christmas, sales and winter collections are in both.
Customers won during the country push are worth more than those won in the same months a year earlier, in every target country. Belgium slightly lower.
Small groups abroad: 46 to 485 customers per bar.
Not what the platforms claim, but what the ads actually add.
Campaigns start when a collection launches, exactly when customers come anyway. Comparing "weeks with ads" to "weeks without" would credit the ads with the launch. So we separate the two.
The figure in our seasonal reports (€7.53 in spring 2026; €16.39 for 2025 including agency fees). Useful for tracking, but it credits the ads with everyone who would have come anyway.
Week by week and country by country, we control for launch demand: store revenue and new Belgian customers (who mostly come without ads) measure how strong each week is anyway.
1. The same collection launch with and without ads (next slide). 2. Spend in other countries does not predict a country's new customers: the effect is not just timing.
The autumn collection launch, 15 Aug – 3 Sep. In 2025 there were no ads anywhere; in 2026 there were, except in Switzerland.
Store revenue was the same both years (€798k vs €794k): the launch pulled equally hard. Yet France went from 22 to 124 new customers, the Netherlands from 5 to 61. Switzerland, without ads in both years: 0 to 3.
One launch, so a snapshot. The figures on the next slide use all campaign weeks of the past year.
Meta + Google spend on all campaigns aimed at that country (acquisition and retargeting).
New customer = first webshop order from that country, in the same week as the spend.
Izoard fees, photo and video production, tools, discounts and free shipping, Global-e and logistics costs. Customers who buy weeks after seeing an ad are not credited, so the true CAC is probably lower.
Every extra €1,000 of Belgian ad spend in a week brings about 11 extra new webshop customers that week, on top of those who come anyway. €1,000 ÷ 11 = roughly €94 per extra customer.
The range is wide (€58 – €263): Belgian demand moves with launches, sales and stores, which makes the ads' share hard to isolate.
Most Belgian customers come through the stores, word of mouth and brand search. Spring 2026 had 40% less spend than spring 2025, and still more new customers.
Belgian media earns its keep differently: retention, launches and store traffic.
Value against cost, country by country, and when to spend.
Gross profit in the first six months per new customer, divided by the incremental CAC. Above 1 = the customer has paid back the media.
Margin to confirm with Buissonnière. Published accounts show €2.88M "gross margin" (Belgian code 9900): revenue minus goods and all services such as rent and logistics, ≈ 31% of revenue. Product margin is higher; we use 50% as a cautious default.
At 50% gross margin, six-month value only. Germany and Switzerland: small samples, wide ranges. Belgium pays back over 18 months (€233), largely via the stores.
Buissonnière is a seasonal business: between launches there is mainly outlet stock of the previous collection. Ads amplify the demand a new collection creates, so that is where the extra budget belongs.
| Country | At the autumn launch 2026 | Average over all campaign weeks | Reading |
|---|---|---|---|
| France | €21 | €48 | Biggest upside: more budget at every launch and capsule |
| Netherlands | €33 | €58 | Same: grow the launch budgets |
| Germany | €37 | €89 | Main launches only |
| Belgium | €28 | ≈ €94 | Launches and retention; the brand does the rest |
Cost per extra new customer. Launch column: 15 Aug – 3 Sep 2026 vs the same weeks of 2025 without ads (one launch, so indicative). The average column covers all weeks with campaigns, launch-controlled.
Three budget scenarios, a recommendation, and a step-by-step scale-up that proves itself as it goes.
Budgets concentrated on launches and capsules, like 2026. Average launch-controlled rates per country; above 1.5× a country's 2026 budget −30% efficiency, above 3× −50%. Not counted: extra purchases by existing customers, the stronger launch effect, late buyers. Cautious figures.
Each scenario pays for itself within 18 months if Buissonnière's gross margin is above a certain level. The extra euros in scenario B need at least 44%; those in C at least 54%.
| # | Move | Why |
|---|---|---|
| 01 | Scenario B: grow abroad | €158k instead of €106k, the extra entirely abroad. Concentrated on the launches and capsule drops, where an extra customer costs ≈ €21–37. |
| 02 | France as the growth engine | Most efficient (€48 on average across campaign weeks, €21 at the autumn launch) and the biggest market. Step up the budget and re-measure each quarter. |
| 03 | Belgium: retention and stores | Brand-driven acquisition, 42% of value in stores. Measure on total value, not webshop ROAS. |
| 04 | Steer on LTV, not ROAS | Confirm the gross margin, then set a maximum CAC per country from 18-month gross profit, reviewed every quarter. |
| 05 | Fix the measurement | Since May 2026, GA4 sees only 6 in 10 orders. Server-side and consent first, or the steering is blind. |
We raise budgets in steps, one country at a time, at the moments a collection launches. Each step shows how many extra customers the next euro buys; the countries not yet stepped up serve as the comparison.
France goes to its scenario-B budget first. NL, DE and CH stay at the 2026 level: they are the comparison.
If France holds its rate at the launch, the Netherlands steps up at the capsule drops. France and NL are compared with DE and CH.
Full scenario B, or C if margin and response allow. Scorecard: back-office new customers per country, not the platforms.
Raising every country at once would mix the effect of the budget with growth and seasonality. Staggering by country keeps a comparison group at every step.
→ Gross margin is an assumption (50%). With the real figure, the unit economics become exact.
→ Incrementality is observational. Launch demand is controlled for and two cross-checks pass, but it is not a randomised test. Scaling in steps, country by country, in 2027 will confirm it.
→ Belgian seasonality is hard to separate from launches and sales periods; the Belgian estimate is the least precise.
→ Germany and Switzerland rest on smaller samples (84 and 79 customers in the push cohort).
→ Values are early: international cohorts are young, so their long-term value is still building.
→ Data: back office to 31 Aug 2026, customers to 3 Sep 2026, ad spend to 30 Sep 2026.