Inverting the real-estate brokerage model
Index qualified demand rather than listings, in a market where 80% of the leads an agency receives are unusable.
- CTO & co-founder
- 2025 — present
- Alacasa SAS
80%
of the leads an independent agency receives are not qualified
Context
The French property market represents roughly one million transactions a year, worth over €250 billion. It counts more than 25,000 agencies, 75% of them independent — one to five people, with a constrained marketing budget.
We target Provence-Alpes-Côte d’Azur and Occitanie: 160,000 annual transactions, €51.5 billion in market value, 4,669 independent agencies.
Every established player — SeLoger, LeBonCoin, PAP — works on the same principle: they index supply. Two and a half million listings online, the same property present on five sites, and no consolidated view of demand anywhere.
Diagnosis
Buyer side
Thirty to fifty listings reviewed before the first viewing, five to eight sites browsed in parallel, three weeks between starting a search and first contacting an agency. The available filters are strictly explicit — price, floor area, town — and structurally incapable of capturing what actually decides a purchase: charm, authenticity, an open view, the character of a place.
Agency side
An independent agency spends on average €680 a month on portal subscriptions. For 80% unqualified leads. Return on investment is very low, and the spend is endured rather than chosen: without the portals, there is no visibility.
Root cause
The two problems are one. The market still indexes supply while scarcity has changed sides: demand is up 13.9% against supply contracting by 2.7%. What is scarce today is no longer the property — it is the serious buyer, with a validated budget and a mature project.
As long as the platform indexes supply, it is selling agencies access to noise it generated itself.
Levers
01 — Invert the index: reference demand
Business benefit: turn an acquisition cost agencies endure into a flow of already-qualified contacts.
Build the first database of qualified buyers rather than one more listings catalogue. The model is B2B2C: buyers create an enriched profile, agencies access contacts whose project is already characterised.
02 — Capture implicit criteria with AI
Business benefit: better match quality, and therefore a conversion rate that justifies the subscription.
The buyer profile combines explicit criteria with implicit preferences detected by the model. This is where the defensible advantage sits: the larger the buyer base grows, the better the understanding of criteria becomes, and the better matches get for everyone.
03 — Pick a defensible niche rather than a broad market
Business benefit: avoid a head-on fight with generalists whose budgets are in another league.
The chosen positioning is “South + atypical”: Provençal mas, bastides, manor houses, character properties. A segment estimated at 8 to 12% of the market, structurally underserved by standardised filters — and where implicit criteria weigh heaviest. Second homes, up 18% since 2020, sit alongside it.
Trajectory
Rollout follows three phases, and their order is not negotiable. A two-sided marketplace cannot be monetised before it has liquidity: charging first is a guaranteed way to end up with neither buyers nor agencies.
- Liquidity — free, to build the qualified buyer base and seed the network effect.
- Light monetisation — introduce a subscription once value is demonstrated on the agency side.
- Maturity — broaden the offer and the geographic footprint.
Where the project stands
The project is in build. At this stage, the deliverables produced are a fifty-page business analysis — market study, segmentation, competitive analysis, business model, go-to-market strategy, risk matrix — and the technical and product architecture of the platform.
My remit covers technology, product and AI. My co-founder Kilian Codaccioni, CEO, leads commercial strategy and business development.
What I took from it
A solid diagnosis is recognisable by what it unifies. Buyer frustration and agency budget waste are the same symptom seen from either side of the market. Treated separately, they yield two half-solutions.
A two-sided model imposes an order. Having the right phases is not enough; they have to run in the right sequence — and that means resisting the pull to monetise early.
Carrying a market intuition all the way down to the database schema changes how you advise. Once you have personally arbitrated between what is strategically desirable and what is technically fundable, you make different recommendations.