At large fleet scale the optimization math does not change, but three other things do. Integration stops being optional, because nobody is pasting lanes into a portal for 500 trucks and the plan has to reach drivers inside the ELD and TMS they already use. Contract leverage becomes real, because a fleet buying eight million gallons a year can negotiate rates a fleet buying one million cannot, and the reporting a vendor gives you becomes a negotiating instrument. And verification stops being a nice-to-have, because at this size a savings claim nobody can audit is a material number in the fuel budget. The arithmetic is simple: at 100,000 miles a year and 6 miles per gallon, 300 trucks burn about 5 million gallons annually, 500 trucks about 8.3 million, and 1,000 trucks about 16.7 million. One cent per gallon is worth roughly $50,000, $83,000 and $167,000 a year at those sizes.
Updated August 2026 · 8-minute read
At 100,000 miles per truck per year and 6 miles per gallon. This is arithmetic, not a projection, so check it against your own fuel spend and mpg. We deliberately do not multiply a savings average out into a headline number for you: apply whatever per-gallon figure you actually believe, including zero, and see what the decision is worth.
| Fleet size | Diesel per year | Value of $0.01/gal | Value of $0.05/gal |
|---|---|---|---|
| 300 trucks | about 5.0 million gallons a year | $50,000 a year | $250,000 a year |
| 500 trucks | about 8.3 million gallons a year | $83,000 a year | $417,000 a year |
| 1,000 trucks | about 16.7 million gallons a year | $167,000 a year | $833,000 a year |
For reference, a blended average of about $0.90 per gallon in reconciled savings across all fleets on Flow (not a per-fleet guarantee; each fleet's result depends on its lanes, contracts, and current fueling habits, and is measured against that fleet's own transactions). Your own number is the one that matters, and it is measured against your transactions rather than estimated. How to audit a savings claim.
The optimization itself is the same problem at every size. What changes is how it has to be wired in, who has to sign off, and what breaks if a human has to watch it.
Where dedicated integration starts paying for itself
Where contract leverage and terminal economics turn on
Where heterogeneity, not scale, is the hard part
These are requirements, not features. A large fleet that cannot get all four should keep shopping, including past us.
At large fleet scale the optimization math does not change, but three other things do. Integration stops being optional, because nobody is pasting lanes into a portal for 500 trucks and the plan has to reach drivers inside the ELD and TMS they already use. Contract leverage becomes real, because a fleet buying eight million gallons a year can negotiate rates a fleet buying one million cannot, and the reporting a vendor gives you becomes a negotiating instrument. And verification stops being a nice-to-have, because at this size a savings claim nobody can audit is a material number in the fuel budget. The arithmetic is simple: at 100,000 miles a year and 6 miles per gallon, 300 trucks burn about 5 million gallons annually, 500 trucks about 8.3 million, and 1,000 trucks about 16.7 million. One cent per gallon is worth roughly $50,000, $83,000 and $167,000 a year at those sizes.
At 100,000 miles per truck per year and 6 miles per gallon, roughly 5.0 million gallons for 300 trucks, 8.3 million for 500, and 16.7 million for 1,000. Adjust for your own mileage and mpg; the useful figure is what a single cent per gallon is worth, which is about $50,000, $83,000 and $167,000 per year at those three sizes. Five cents is about $250,000, $417,000 and $833,000. Those are arithmetic, not projections, which is why they are worth more than a vendor's savings estimate: you can check them against your own fuel spend in a minute.
It should be, and it should be scoped rather than list price. Flow is per truck per month with Enterprise from $49, scoped through a demo because at this size the work is the integration: which ELDs, which TMS, how many fuel card feeds, whether terminal fills are in play, and how the rollout is phased. Any vendor quoting a large fleet from a pricing page without asking those questions has not understood the job.
It is the normal case at this size and it should not be a problem. Flow's integrations are configured per tenant rather than assumed, and more than one active ELD for a single fleet is a supported configuration rather than an edge case. What matters in evaluation is asking the vendor directly how they handle it, because a tool built around one ELD per carrier will quietly mis-plan for the trucks on the other one.
Yes, and you should. The normal path is one terminal, region or division first. Just as important, plans can generate and stay visible in the portal while nothing at all is sent to drivers, so operations can judge plan quality on real loads before a single message reaches a cab. Turning driver-facing delivery on is the carrier's decision and stays under the carrier's control.
Flow runs across about 1,000 trucks today, spread over five fleets on different ELDs, fuel platforms, and driver communication channels, including a carrier in the 300 truck range. That heterogeneity is the relevant part: the hard problem at large fleet scale is rarely volume, it is that no two carriers have the same stack. We are happy to talk through how a fleet of your size and shape would be integrated, in specifics, during a demo.
Put the same questions to everyone: is each recommendation reconciled against the fuel card transaction that followed it, is price prediction accuracy published as a number, what happens to a plan when dispatch reroutes a truck mid-day, and which states or chains is the pricing data weakest in. Our full walkthrough of how to audit a savings claim is at /guides/fuel-savings-you-can-verify, and it applies to us as much as to anyone else.
Related: how to audit a savings claim, the category, compared, or how integrations work.
Bring your ELDs, your TMS, your fuel card feeds and a few real lanes. A 30-minute demo shows the plans on your own routes and the reconciliation trail behind them, before anything reaches a driver.