French AI Startup Actionable Raises $10m to Predict Customer Behaviour
Paris-based Actionable, founded by ex-Criteo Europe boss Nicolas Rieul, has raised $10m to scale predictive and causal AI that scores entire customer journeys.
The raise
Actionable, a Paris martech startup, has raised $10m in a round led by Hi inov with Axeleo Capital and more than 30 tech investors. Founded in April 2024 by Nicolas Rieul (former CEO for Europe at Criteo) and Nans Thomas (ex-CPO of Innovorder), it says it already manages 117 million consumer journeys for 16 major clients including Carrefour, OUIGO and Engie.
The platform scores churn risk, satisfaction, complaint severity and repurchase likelihood across the customer journey. The new funds, announced on 7 September 2026, are earmarked for product, engineering and sales hiring, plus US expansion via resellers.
Beyond attribution, toward prediction
The pitch is a deliberate step past attribution. Instead of explaining what already happened, Actionable's causal-AI approach tries to predict what a specific customer will do next — and, crucially, what intervention changes it. That framing is resonating with big French retailers and transport brands hungry for retention gains.
Our read: the promise is seductive and the client logos are real, but marketers should hold the claims to a hard standard. "Predictive" and "causal" are among the most over-used words in martech. The question is not whether the model outputs a churn score, but whether acting on it demonstrably moves revenue.
How to test tools like this
First, insist on a controlled pilot: hold out a control group, act on the model's scores for the test group, and measure the incremental difference in retention or repurchase. A causal claim that cannot survive a holdout test is just a correlation with better branding.
Second, watch the operational cost. Predictive scores only pay off if your teams can act on them at speed — the value is in the intervention, not the dashboard. Before rewiring budgets around a tool like this, be sure the organisation can actually do something with what it predicts.
Quick FAQ
How much did Actionable raise and from whom?
$10m, led by Hi inov with Axeleo Capital and 30-plus tech investors, announced on 7 September 2026.
What does Actionable's platform do?
It uses predictive and causal AI to score churn risk, satisfaction, complaint severity and repurchase likelihood across the customer journey, for clients including Carrefour, OUIGO and Engie.
How should marketers evaluate it?
Run a controlled pilot with a holdout group and measure whether acting on the scores demonstrably lifts retention or revenue.
Every martech tool now promises to be 'predictive' and 'causal' — the words are almost meaningless. The only test that matters is a holdout: act on the scores for one group, not the other, and see if revenue actually moves. If a causal claim can't survive that, it's a correlation in a nicer suit. — Daniel Nikolla, Founder of Merx Marketing
Predictive martech earns its budget in a holdout test, not a dashboard — measure the incremental revenue from acting on the scores before you rewire anything.
When a vendor says its AI is 'causal', can you run the holdout test that would actually prove it?
If you would like to know more about this topic, please contact us on danieln@merxmarketing.co.uk
Sources: e-marketing.fr, CB News.
Related reading
Written by Daniel Nikolla, Founder of Merx Marketing Ltd and Marketing Minute
Comments