The HVAC units that fail in July are already showing it in May. Here's how to find them before the truck rolls start.
Every multi-site operator sees the same pattern. April runs clean. May gets choppy. By the second week of July, the HVAC ticket queue is full and emergency-rate service trucks are out across the portfolio. Then the spoilage emails start.
This is not new information. What's new is the math on what it costs in 2026.
A small to mid-size restaurant with one refrigeration failure runs $3,000 to $15,000 in spoilage per incident. High-volume operations push past $50,000. A single 4-hour failure in a 20,000 square foot supermarket can total $50,000 of inventory loss. Health regulations require disposal once food sits above 40 degrees for four hours, so the call is binary and the clock is short.
Add the truck-roll cost on top. Emergency-rate HVAC service in summer runs 1.5 to 2x standard rates. The techs who can get to your site at 11pm on a Friday are the ones who already booked their time at premium pricing.
This is why summer ops budgets blow up. Not because of the volume of failures. Because of when they happen and how they cluster.
Most multi-site operators have the data to predict July's failure list right now. They just don't read it that way.
Every HVAC unit running in March and April produces a thermal and electrical signature. Units already drawing more current than their nameplate, cycling more frequently than peers in the same region, or struggling to hold setpoint on warmer days are telling you exactly what's going to happen when ambient temps hit 95 degrees in July.
The ones that are degraded but not failed are the most expensive failures, because they fail the latest, in the worst conditions, on the busiest nights.
Reading the early signal is not a moonshot. It's a triage exercise across your portfolio. Sort sites by which units are already showing strain, prioritize service before peak season, and the entire summer cost curve compresses.
The 60-day window to do this closes on June 30.
The National Restaurant Association says industry sales hit $1.55 trillion in 2026. It also says 42 percent of operators are not profitable. Food costs sit 35 percent above pre-pandemic. Tariffs hit 68 percent of operators. Labor is a median 36.5 percent of full-service sales.
Pricing is maxed out. Labor is compressed. The lever still on the table is operating cost. For multi-site groups, equipment uptime and energy are two of the few categories where you can take 8 to 10 percent out without touching the customer experience.
GlacierGrid's portfolio runs 8.12 percent energy savings across active customers as of Q2 2026. The savings did not come from anyone changing how they ran their stores. They came from catching equipment behavior earlier and acting on it.
For multi-site operators with 50 or more locations, the pre-summer move is four steps:
The operators who do this in May spend service-truck dollars at standard rates. The operators who skip it spend them at emergency rates in July and pay spoilage on top.
Most of the cost difference between a clean summer and a punishing one is not on the maintenance line. It's on the spoilage and emergency-service lines, in the months when the calendar is least forgiving.
GlacierGrid just shipped Cooling Readiness reports to active customers, covering March 17 through April 16 data per site. Each report identifies which HVAC units are showing strain and ranks them by likelihood of summer failure. The methodology splits SAT and non-SAT sites because the leading indicators differ.
Operators who don't run on GlacierGrid can do a version of this with their own data. The five indicators that matter most:
Run those across the portfolio. The list ranks itself.
Today is May 4. The window to act before peak summer is 60 days.
If you run 50 or more locations, the question is not whether some HVAC units will fail in July. The question is whether you find them in May at standard rates, or in July at emergency rates with spoilage attached.
GlacierGrid offers a free 90-day pilot for operators who want a portfolio-wide read. Bring your March and April HVAC data, get a cooling readiness scorecard back, and decide before June whether the visibility is worth keeping.