"Retail IoT" gets used to mean three different things, and the content out there often mixes them. Supply-chain visibility (where the truck is, where the pallet is). Store-floor analytics (foot traffic, dwell time, queue length). And facilities monitoring (HVAC, refrigeration, energy, lighting). Three categories, three buyers, three software stacks.
This is about the third one. Retail IoT platform must-haves for facilities teams running 50 to 500 stores. If you're sourcing for supply chain or floor analytics, the rest of this won't help.
The eight capabilities below are the ones that separate platforms that survive a fleet rollout from the ones that get quietly shelved by month six.
Multi-site retail HVAC runs on legacy controls already in the building. Honeywell, Carrier, Trane, Daikin, plus a long tail from acquisitions. The platform has to ingest data from what's installed today, not require a controls swap before any value flows.
Ask the vendor: "Here are the BAS vendors across our fleet. What does integration look like for each?" If the answer involves a months-long controls project before the dashboard lights up, the platform is the wrong category for a multi-site rollout.
A facilities director needs the fleet view in one screen: sites trending off baseline, sites with active alerts, sites stable. A facilities engineer needs the drill-down to a specific RTU at a specific store, with the actual sensor curve visible.
Both views matter. A platform that ships only the rollup loses the engineer. A platform that ships only the unit data loses the director.
Ask the vendor: "Show me the dashboard with 200 sites loaded. How long does it take to find the three worst-drifting RTUs this week, then drill into the actual amp draw curve on one of them?"
Alerting is where platforms quietly lose their users. If half the alerts are false positives, the team starts ignoring all of them. By month four, the platform is open zero times per day.
Good alerting suppresses door-open events as anomalies, distinguishes scheduled defrost from drift, accepts a holiday as not-a-Tuesday, and doesn't fire on every two-degree fluctuation.
Ask the vendor: "What's the false-positive rate on your alerts in real customer deployments? How does it tune over time?"
The alert is half the work. The other half is turning the alert into a work order assigned to the right vendor with the right context, automatically.
A platform that flags an issue and stops there forces the facilities team to be a translation layer. The team won't do that for long. The platform has to push directly into the CMMS or service-management system the chain already runs.
Ask the vendor: "Show me the integration with our CMMS. What fields populate automatically, what requires a human to fill in?"
Multi-site retailers run BI on top of operational data. They merge facilities data with energy spend, with same-store sales, with weather. A platform that owns the data and rents it back via a clunky export is dead weight in that stack.
The chain should own its data. Full API access. CSV export when the API isn't enough. No vendor lock on the operator's own equipment data.
Ask the vendor: "What's the API documentation look like? Can we pipe data into our own warehouse on a schedule?"
Every setpoint change, every override, every alert acknowledgment, every work-order status change. Time-stamped. User-attributed. Queryable.
The audit trail is non-negotiable for two reasons. First, energy-savings claims get challenged in budget review. The audit log is what holds the claim up. Second, HACCP and food-safety compliance requires temperature audit records that a regulator can verify.
Ask the vendor: "Show me the audit log. How far back does it go? What's the export format?"
Store WiFi is the worst data path in retail. It goes down. It gets rebooted. It gets put on a guest network with throttling. A facilities platform that depends on store WiFi loses data exactly when it's needed most: during outages, during peak load, during the failure event.
The data layer needs cellular or LoRaWAN backhaul, ideally with a fallback path. The hardware should keep buffering data locally when the link drops, and sync when it comes back.
Ask the vendor: "What happens when the gateway loses connectivity for 12 hours? Where's the data, and when do alerts fire after reconnect?"
The 12-month enterprise rollout is how multi-site IT projects fail. By the time the platform is fully deployed, the operator has spent the budget, lost the executive sponsor, and has nothing to show.
A facilities platform that's worth deploying can prove value in 90 days on a 10-store pilot. Real RTUs. Real data. Auditable savings. ~10% energy savings, 1-month payback on the platform itself, and 15% fewer service calls are realistic pilot benchmarks.
Ask the vendor: "What does a 90-day pilot on 10 stores look like? What numbers do you commit to producing?"
Walk it through every vendor on the shortlist. Not as a yes/no scorecard. As an open-ended question framework that forces specifics.
Vendors who can't answer one or two cleanly aren't disqualified. Vendors who can't answer five aren't deployed at multi-site facilities scale, regardless of what the website says.
GlacierGrid HVAC Intelligence is the multi-site intelligent platform for retail, c-store, and restaurant operators running 50 to 500 stores. Designed against this exact checklist. Run the eight questions against any vendor, including GlacierGrid, and pick the one that answers cleanly.
Start a free pilot: 90 days, real stores, real data, no long-term contract. Learn more about GlacierGrid HVAC Intelligence.