From €20/lead to under €3.
A French manufacturer of glass garde-corps and minimalist fenêtres. Two product lines, two markets — B2B (architects, contractors) and B2C (homeowners). 284 qualified leads at €3.57 CPL with a 3D configurator that pre-qualifies every prospect. And an 85% CPL reduction on the B2B line — through disciplined testing, not paid magic.
A French manufacturer of glass garde-corps (railings, balustrades) and minimalist fenêtres (windows), doors, and sliding doors. Two distinct product lines. Two distinct markets: B2B (architects, contractors, developers) and B2C (homeowners renovating or building).
The owner came to us for Meta Ads on his core B2B product line — minimalist windows, doors, and sliding doors for professionals. Glass railings were a secondary product with untapped potential.
Then he saw something at a major competitor's site: a 3D configurator that let customers design their own railing setup online. He asked if we could build one. We said yes — but not before we understood every detail of the product.
What we built — step by step
- 01
Product deep-dive
Sessions with the owner to understand every configuration option — glass types, mounting styles, dimensions, finishes, accessories. Every detail that affects pricing, feasibility, and buyer intent. We didn't just build a configurator. We built a lead qualification machine.
- 02
Market research
Before creating a single ad, we researched the French garde-corps market. Pains, desires, objections, buying triggers — pulled from forums, competitor reviews, and industry data. We found what homeowners actually worry about (safety, aesthetics, price transparency) and what contractors care about (specs, lead time, bulk pricing).
- 03
Two landing pages — B2B and B2C
One page for homeowners who want to visualize their project and get a quote. One page for professionals who need specs, pricing, and fast turnaround. Same product, completely different language, completely different conversion path.
- 04
Persona mapping and ad angles
Based on the research, we mapped the personas most likely to buy and created specific angles for each: the homeowner renovating a terrace, the architect specifying materials for a new build, the contractor looking for a reliable supplier. Each persona got its own creative angle.
- 05
3D Configurator — the conversion engine
The configurator lets visitors design their own garde-corps project: choose glass type, railing style, dimensions, accessories. Every completed configuration is a submitted application with specific requirements — not a generic 'contact us' form fill. The prospect has already invested time, made decisions, and seen a visual of their project. They're pre-sold.
- 06
Meta Ads — launch and optimization
Campaign structure built around the two audiences (B2B and B2C), with creative angles from the research. 9 campaigns tested across audiences, angles, and formats. Weekly optimization, creative rotation, audience refinement.
How we cut CPL by 85% on windows, doors & sliding doors
- 01
Round 1 — Research
Before touching a single campaign, we went back to fundamentals. Who's actually buying? What language do they use? What triggers the purchase decision? What do competitors say — and what do they miss? We rebuilt the creative strategy from scratch based on real data.
- 02
Round 2 — Creative testing
Multiple ad variations — different hooks, different visuals, different angles. Not 'let's try blue vs red.' Real creative hypotheses: does a contractor respond to 'faster installation' or 'higher margin per project'? Does an architect care about 'design flexibility' or 'client satisfaction scores'? We tested until we found the angles that moved the needle.
- 03
Round 3 — City testing
France isn't one market — it's dozens. A contractor in Marseille doesn't respond to the same ad as one in Lyon. We ran geo-targeted campaigns across cities and regions, identified which markets converted at the lowest cost, and reallocated budget accordingly. Some cities delivered €2-3 leads. Others were €15+. We killed the expensive ones and doubled down on what worked.
- 04
Round 4 — Age testing
Not all contractors are the same age, and age correlates with buying behavior. Younger contractors respond to different messaging than established firms. We segmented by age brackets, tested creative against each, and found the sweet spots where CPL dropped dramatically.
- 05
Round 5 — Scale what works
After four rounds — research, creative, geo, demographics — we had a system: the right message, to the right person, in the right city, at the right age bracket. CPL dropped from €20 to under €5-7 consistently. On the best periods: €2-3 per lead. Same budget. Same product. Completely different results.
This wasn't one magic trick. It was months of systematic, disciplined testing. One variable at a time. Each round of tests feeding the next.
The numbers
€2.15–€3.57 leads are people who expressed interest — clicked, filled a form, want to know more. Volume plays. Good for pipeline building.
55 submitted applications at ~€13 each are different. These are prospects who opened the 3D configurator, designed their project, selected their materials, and submitted a detailed request. They already know what they want. They've already visualized the product in their space. The sales team's job is to close — not to educate.
How CPL dropped 85%
That's an 85% reduction in cost per lead. Not by increasing budget. Not by changing the product. By doing the work most agencies skip: research, systematic testing, and disciplined optimization.
Across all product lines
Why most agencies can't do this
This took months of work. Four rounds of testing. Hundreds of ad variations. Weekly analysis. Constant creative rotation. Most agencies don't have the patience or the process for this. They launch 3 ads, watch for a week, and say 'the algorithm needs more data.' That's not optimization. That's laziness.
What we did was build a testing protocol: one variable at a time, documented results, clear kill criteria, and a rule — never scale before you've found the floor on CPL. It's not glamorous. It works.
All garde-corps metrics from Meta Ads Manager (9 campaigns). B2B optimization journey documented across multiple testing rounds. €5,633 total client spend across all product lines. No numbers rounded up, no metrics cherry-picked.