This French AI Startup Just Raised $10M to Predict Which Customers Will Leave

Jan de Vries · 2026-09-09

You know that sinking feeling when a customer churns without warning? A Paris-based startup called Actionable thinks it has the answer, and it just raised $10 million to prove it. The company uses artificial intelligence to predict which of your customers are about to leave, complain, or buy again—and, crucially, it can tell you why. That kind of foresight is gold for large enterprises drowning in customer data but starving for actionable insights. ### The Funding Round Actionable announced a $10 million round led by Hi inov, with pre-seed investor Axeleo Capital increasing its stake. This follows a $2.3 million pre-Seed round announced in September 2024. Not bad for a company founded just last year. "Value is moving from collecting data to intelligence," said Valérie Gombart, co-founder and CEO of Hi inov. "Enterprises still pay heavily to store and measure data, when what is missing is the layer that explains and proposes a plan of action, for teams and for AI systems alike." ### The Problem They're Solving Co-founders Nans Thomas and Nicolas Rieul have been around the block. Rieul held general management roles at Criteo and chaired Alliance Digitale, the French digital industry association. Thomas was CPO at Innovorder and founded Wino, a tech platform for local retailers. Rieul kept seeing the same frustrating scenario across large French companies: teams collecting Net Promoter Scores and churn data, but nobody able to act on it. "For fifteen years, I sat in the same Monday morning meeting," Rieul said. "A dashboard says satisfaction or revenue is down two points, everyone around the table looks for an explanation, and nobody can say which customers to call the next day or what to decide right now." That gap between measurement and decision-making is exactly what Actionable targets. ### How the Platform Works When a client sends raw tabular data, Actionable automatically rebuilds the customer journey inside what it calls a "Common Customer Data Model." This model is tailored to specific industries—retail, financial services, insurance, transport, energy, telecoms, and automotive—and incorporates operational realities like waiting times, loyalty cards, pricing systems, load factors, and delivery schedules. "Putting an LLM on top of a data warehouse is not enough," Thomas explained. "Without business context, an AI reads raw tables very badly. The hard part is turning hundreds of tables and in-house definitions into a customer model a machine can use without getting it wrong." ### Real Results for Big Names Actionable claims to map the journeys of 117 million consumers for large enterprises. Notable clients include Carrefour, the French state railway SNCF, utility Engie, and employee-benefits group Edenred. At OUIGO, SNCF's low-cost rail brand, the platform predicts passenger satisfaction daily. Here's what makes the approach different: - It works with 100% of the customer base, not just a sample - Every customer gets scores for churn, satisfaction, serious-complaint risk, and repeat purchase - Each score comes with the cause attached, so teams know exactly what to fix - Its Actionable Intelligence agent can produce analyses in hours that would take a human analyst weeks ### Why This Matters Now Companies have spent years collecting customer data, but most still struggle to turn it into decisions. Actionable's bet is that the next competitive edge isn't about having more data—it's about having the intelligence layer that explains what it means and suggests what to do next. As AI agents become more common in CRM and customer service, having clean, contextual customer data will only become more critical. Actionable is positioning itself as the infrastructure those agents need to actually be useful. For US-based professionals watching the European tech scene, this is a startup worth tracking. The playbook of combining AI with deep industry context could easily translate across the Atlantic.