Ask for the location count.
That single question separates the vendors in multi-site HVAC management faster than any feature comparison, and it works on us too. A 14% savings figure means one thing across four stores and something entirely different across four hundred. Percentages are easy to produce on a small, favorable sample. They get much harder to hold as the footprint grows, because a portfolio at scale includes the difficult sites: the old rooftop units, the store where the manager props the back door, the location whose utility rate structure punishes exactly the hours you are open.
So when you evaluate a platform, ask three things about every number on the page:
A vendor confident in their results will answer all three without hesitating. Here are ours.
| Operator | Result | Locations |
|---|---|---|
| A QSR franchise group | $1.1M annual savings | 181 locations |
| A QSR franchise group | $34,000 per month | 70 locations |
| A regional coffee chain | 10% energy reduction | 106 locations |
| A QSR franchise group | $250,000 annual savings | 38 locations |
| A multi-unit QSR operator | 12.8% energy reduction | Multi-location portfolio |
These are measured results from live portfolios, not modeled projections. Operator names are withheld; we publish the location count and the measurement basis because those are the parts you can actually evaluate.
The reason we lead with location counts rather than the biggest percentage is that the percentage is the easy half. Holding roughly 10% across 181 locations is a materially harder thing to do than hitting 24% at a handful of sites, and if you operate at scale you already know why.
You will see the claim that operational waste runs 5% to 10% of energy spend and can be recovered with no capital investment. That is broadly right, and we would not argue with it.
What it leaves out is who does the recovering.
Operational waste is recoverable without capital because the fix is behavioral and schedule-driven rather than equipment-driven: setpoints that drift, schedules that never got updated after hours changed, units fighting each other across a shared space, doors left open on refrigeration. None of that requires new hardware to correct. All of it requires someone to notice, decide, and act, at every location, continuously.
At one site a good facilities manager does this by walking the building. At fifty sites nobody walks the building. The waste is recoverable in principle and unrecovered in practice, which is why it is still sitting there to be claimed as a headline.
The question to put to any platform making this claim is whether it tells you about the drift or corrects it. Monitoring produces a report someone has to action. Control changes the setpoint. Across a large portfolio that difference compounds every day.
To be precise about what we do, because the distinction matters: GlacierGrid controls HVAC and lighting. We monitor refrigeration, and alert on it before product is lost. We do not claim leak detection.
Portfolio energy benchmarking is genuinely useful. Ranking every location by energy intensity surfaces your worst performers, and any operator running more than a few dozen sites should be doing it.
It also stops exactly there.
A benchmark tells you that store 114 is in the bottom decile. It does not tell you whether that is a failing compressor, a schedule nobody updated after the hours changed, a rate structure that makes the same kilowatt-hour cost more, or a store that is simply busier than its neighbors and should be using more energy. Those four causes need four different responses, and three of them are not the store manager's fault.
Ranking creates accountability. Diagnosis creates savings. If your platform gives you a leaderboard and leaves the diagnosis to your facilities team, you have added a reporting obligation rather than removed a workload.
Run it on your own locations first.
GlacierGrid's pilot is 90 days, free, on your sites, measuring your savings against your baseline. No purchase, no commitment to continue.
We offer it because every argument above cuts against us too. You should not take our 181-location number on faith any more than you should take anyone else's projection on faith. The only number that settles the question is the one produced by your own portfolio, on your own equipment, in your own rate territory, over a period long enough to survive a weather swing.
If a vendor is confident their platform performs, a proof period costs them nothing. Notice which ones offer it.
Bring these to every conversation in this category, including ours:
Any vendor worth shortlisting will answer all six plainly.