No vendor can honestly tell you it saves the most per gallon, because savings depend on your lanes, your existing contracts, and where your trucks fuel today. The question that can be answered is which vendor will prove what it saved. Ask two things of every vendor on your shortlist: is each recommendation matched to the actual fuel card transaction that followed it, showing the stop taken, the miles off plan, and the net dollars gained or lost; and is price prediction accuracy published as a number, in cents per gallon, against real pump prices. A vendor that does both has given you an audit trail. A vendor that does neither is asking you to accept a figure it calculated about its own performance.
Updated August 2026 · 7-minute read
These apply to every vendor in the category, us included. If a vendor cannot answer one of them in a demo, that is itself the answer.
The one that matters most
The honesty test
Where most savings quietly leak
Check the incentive
Nobody volunteers this
We wrote this guide, so the fair thing is to answer our own questions first. We are not claiming other vendors cannot do these things. We are showing what we put in writing, and inviting you to ask everyone on your shortlist for the same.
The wider field, categorized honestly: best fuel optimization software, Trimble Expert Fuel alternatives, or Fuel Router alternatives.
No vendor can honestly tell you it saves the most per gallon, because savings depend on your lanes, your existing contracts, and where your trucks fuel today. The question that can be answered is which vendor will prove what it saved. Ask two things of every vendor on your shortlist: is each recommendation matched to the actual fuel card transaction that followed it, showing the stop taken, the miles off plan, and the net dollars gained or lost; and is price prediction accuracy published as a number, in cents per gallon, against real pump prices. A vendor that does both has given you an audit trail. A vendor that does neither is asking you to accept a figure it calculated about its own performance.
Because the number is mostly a property of your fleet, not the software. A carrier already holding strong Pilot, Love's and TA contracts whose drivers mostly fuel at the right stops has less to recover than one with good contracts and no enforcement. Lane mix matters too: a fleet crossing high-tax and low-tax states repeatedly has more IFTA arbitrage available than one running inside a single state. Any vendor quoting a guaranteed cents-per-gallon figure before seeing your lanes and your contracts is quoting a marketing number.
It means every dollar of claimed savings is tied back to a real fuel card transaction. For each fill, you see the stop the software recommended, the stop the driver actually used, how far off plan that was, and the net dollars gained or lost after discounts and state fuel tax. Without it, a savings figure is the vendor's own model of its own performance, with nobody checking. With it, you can click into any single trip and audit the claim. This is the difference between a reported number and a verifiable one.
Ask for the number rather than a standard. What matters is that the vendor measures predicted price against the actual pump price and publishes the result: mean error in cents per gallon, and the share of predictions landing within 5 cents and within 10 cents. Flow publishes this in the dashboard. Any vendor can compute it, since it only requires comparing their own recommendation to the transaction that followed, so the useful signal is whether they show it to you unprompted.
They prove what was spent, not what was saved. Card vendors report savings calculated by the vendor against a baseline the vendor chooses, with no independent verification. That is not dishonesty, it is just a different job: a card is a payment instrument with controls and a discount schedule attached. Deciding where a truck should have fueled, and then checking whether it did, is what a route-aware optimizer is for. The two work together, and the reconciliation question is the one worth asking of the optimizer.
Four questions. Show me one trip end to end, from the recommendation to the transaction that followed it. What is your price prediction error in cents per gallon in my operating region. What happens to a plan when dispatch reroutes the truck mid-day. And which states or truck stop chains is your pricing data weakest in. Vendors who can answer all four quickly, including the uncomfortable last one, are the short list. Put these to us as well.
A 30-minute demo runs your own lanes and shows the reconciliation trail on a real trip, including the fills where a driver beat our recommendation. Then ask every other vendor the same five.