A negotiated discount is only worth what gets used, and most fleets lose the gap between the two because nothing checks it load by load. The rate exists in a contract, the driver decides where to stop, and no system compares the two until the invoice arrives. Closing it takes four things: load every contract you hold into one place so the true net price at each stop can be compared, decide the stop per load rather than per driver habit, deliver that decision inside the ELD or SMS the driver already reads so following it is easier than not, and reconcile every transaction afterwards against the recommendation so you can see which loads complied and what the misses cost. None of that requires changing fuel cards or joining a fuel network. The discounts you already negotiated are usually enough; they are just not landing.
Updated August 2026 · 7-minute read
Every one of these is a normal, well-run fleet. None of them require a bad driver or a bad contract, which is why the leak is so easy to miss on a fuel invoice.
Leak 1
Leak 2
Leak 3
Leak 4
In order. Skipping straight to enforcement without the first two is why discount programs stall.
A negotiated discount is only worth what gets used, and most fleets lose the gap between the two because nothing checks it load by load. The rate exists in a contract, the driver decides where to stop, and no system compares the two until the invoice arrives. Closing it takes four things: load every contract you hold into one place so the true net price at each stop can be compared, decide the stop per load rather than per driver habit, deliver that decision inside the ELD or SMS the driver already reads so following it is easier than not, and reconcile every transaction afterwards against the recommendation so you can see which loads complied and what the misses cost. None of that requires changing fuel cards or joining a fuel network. The discounts you already negotiated are usually enough; they are just not landing.
No, and be skeptical of anyone who says otherwise. The problem is almost never the card or the rate, it is that nothing translates the rate into a per-load decision and then checks whether it was followed. A route-aware optimizer sits on top of the cards and contracts you already hold. If a vendor's answer to poor discount capture is to move your volume onto their network, they are solving their revenue problem rather than yours.
Usually for reasons that are not about fuel: parking availability, showers, a familiar exit, or a plan that arrived too late to act on. Treating it as a discipline problem tends to backfire. The more effective approach is to remove the friction, by putting one clear recommendation with ranked backups inside the ELD message the driver already reads, and then to measure. Flow scores off-plan stops in both directions, so the ones where a driver found a better deal than the plan show up as a data quality signal on our side rather than a mark against the driver.
You can bound it from your own data before buying anything. Take a month of fuel transactions, and for each one ask what the net price would have been at the best contracted stop within a reasonable detour on that route. The difference is the leak. Most fleets have never run that comparison because it requires the route, the tank state, and the full contract set together. That comparison is exactly what per-transaction reconciliation automates, which is why it is worth asking every vendor whether they do it.
Yes. Plans can reach a driver by SMS or an emailed link with no ELD involvement at all, which is how owner-operators and mixed fleets run. ELD in-cab delivery is better when it is available because it puts the plan next to the dispatch the driver is already reading, but it is not a requirement.
It will probably disagree with it, and that is useful. Card vendors report savings they calculate themselves against a baseline they choose, with no independent verification. Per-transaction reconciliation compares each actual purchase against a specific alternative that was available on that route at that time. When the two numbers differ, the reconciled one is the one you can audit line by line.
Related: how to audit a savings claim, the category, compared, how integrations work.
A 30-minute demo runs your real lanes and contracts through Flow and shows, load by load, where the discount you negotiated did and did not land.