Used Car Listing Collection at Scale
French automotive analytics startup
We replaced a cost-prohibitive third-party data provider with a custom scraping pipeline that collects 9 million+ used car listings per month at 85% lower cost. The client went from data dependency to full ownership of their core asset.
The Challenge
The client’s business model depended on large-scale used car listing data, but their third-party provider charged roughly € 0.10 per listing. At the volumes needed to cover the market, the cost was unsustainable and capped the company’s growth.
The client sells market intelligence to car dealerships, insurers, and used car sellers. Their product depends on comprehensive, fresh listing data across the entire French used car market. At roughly € 0.10 per listing from external providers, scaling to full market coverage was economically impossible. The cost structure had been a constraint since the company’s founding, limiting both the depth and breadth of insights they could offer clients.
Our Approach
We built a custom scraping pipeline with proxy rotation and resilient extraction logic, designed to collect listings at scale from every major used car platform in France. The pipeline runs daily and feeds directly into the client’s analytics infrastructure. The project started with 3 sites and has expanded to cover the entire sector.
What We Built
Full-market coverage
Scraping agents collect from every major used car listing platform in France, giving the client exhaustive market visibility.
Proxy rotation and resilience
Built to handle anti-scraping protections at scale, with automatic recovery from blocks and layout changes.
Daily collection
Listings collected every day, keeping the client’s analytics current with the market.
Ongoing expansion
Started with 3 sites, now covers every relevant platform in the sector. New sources added as the market evolves.
Results
The client went from constrained sampling to full market coverage across every major platform in France. Data costs dropped to a fraction of the previous model, removing the economic ceiling that had limited the product since founding. The shift to an in-house pipeline gave the client direct control over data quality and collection priorities.
Before & After
| Metric | Before | After |
|---|---|---|
| Monthly listings collected | 1 million | 9 million+ |
| Cost per listing | ~€ 0.10 (third-party) | 85% lower |
| Data pipeline ownership | Third-party dependency | Full in-house control |
| Platform coverage | 3 sites | Full sector coverage |
| Time to first results | --- | 4 weeks |