+12.4% gross margin. From 1 to 3 cents.
Causal AI for fuel retail pricing — proven in a live 2-year pilot across a 4-station network.
4-station pilot, 110-week backtest. Indicative of opportunity, not a guarantee of future results.
Most fuel pricing tools tell you what your competitors charge.
None of them can tell you what your own price move does to your margin. That gap costs operators money on every pricing cycle — silently, at scale.
4-station network, 110-week backtest. Indicative of opportunity, not a guarantee of future performance.
Three steps. One daily recommendation.
Measure your actual price sensitivity
Double Machine Learning isolates how your own customers respond to price — per station, per fuel type — from your existing data. No live experiments. No guesswork.
Prove it before acting on it
Every sensitivity estimate is tested for statistical significance before it is used. If the data cannot prove the effect is real, the recommendation is held back. Seven of eight product lines in our pilot cleared the bar. One did not — and was not priced.
Recommend safely — with an explanation
A daily price per station arrives with a plain-language reason: why this price, what the competitive position looks like, what the expected trade-off is. The manager reviews and approves. Nothing goes live automatically.
Read the full methodology and pilot results
The complete breakdown: who was in the pilot, how gross margin was measured, the per-station statistical evidence, and what the 12.4% actually means in practice.
Your details are used only to send the study. No spam, no sequences without your consent.
Prefer to talk first? Book a demo instead
Double Machine Learning — peer-reviewed causal inference. Chernozhukov et al., 2018, The Econometrics Journal.
Minimal risk classification under the EU AI Act. Explainable outputs, anonymised data, on-premise deployment available.
Manager approval on every recommendation. Hard guardrails. Never below cost. Never outside proven range.
See it on a network like yours.
Book a short demo and we will show you how Pricing Intelligence works with your stations, your competitors, your market.