Retail Connectivity: What One Hour of POS Downtime Actually Costs
When a register goes dark, the first thing everyone does is stare at the card reader. Then the manager walks over. Then a customer puts down a full cart and walks out.
That one scene is why retail connectivity is a revenue problem, not an IT problem.
The number nobody budgets for
Most retailers can quote their average transaction to the penny. Almost none can tell you what a minute of downtime costs across the fleet.
Do the math on a store doing $2 million a year. That's roughly $5,500 a day, about $460 an hour across a twelve-hour day. Now multiply by the number of registers that go dark, the number of stores, and how long the outage lasts.
An hour-long outage at a single store is a rounding error. The same outage across 40 stores is a six-figure day. And it happens more than anyone wants to admit — a carrier hiccup here, a failed modem there, a firmware update that takes a router down at 11 a.m.
What actually breaks
Retail networks are fragile because they're bolted together over time. A store that opened with two phone lines now runs:
- POS and payment terminals
- Inventory and price management
- Loyalty programs and gift cards
- Security cameras and alarms
- VoIP phones and paging
- Wi-Fi for staff and sometimes customers
All of it rides on a single broadband connection that was sized for 2016.
The failure is rarely the POS software. It's the last mile. The cable modem loses sync. The DSL line degrades. The router overheats in a back room with no ventilation. And because there's no failover, one dead link takes every system in the store down with it.
The cost that doesn't show up on the P&L
Lost sales are the headline, but they're not the whole story.
Abandoned carts. A customer who can't check out doesn't just wait. They leave. And a chunk of them don't come back.
Refunds and reconciliation. When a card is charged but the receipt never prints, or a transaction posts twice, someone has to untangle it. That's labor.
Inventory drift. Sales that don't ring up don't update stock counts. Next week you're showing 14 units on hand when the shelf is empty — and losing the sale all over again.
The manager's time. Every outage turns a store manager into a network technician, on the phone with a carrier instead of on the floor.
None of those show up under "IT." They show up as soft margin erosion that's hard to attribute and easy to ignore.
The fix is failover, not faster internet
The instinct is to buy more bandwidth. That doesn't help when the whole line is down.
What actually protects a store is a second path. A primary connection with a separate failover on a different carrier — cellular or fixed wireless — that cuts over automatically. The register never knows anything happened. The customer never knows either.
Here's the part most retailers get wrong: failover only matters if it's tested. A backup that hasn't been flipped in six months isn't a backup. It's a wish.
Build it once, roll it out everywhere
For a single store, this is a solved problem. The hard part is scale.
A retail chain with dozens or hundreds of locations doesn't need a unique engineering project per store. It needs a standard: same hardware, same failover, same configuration, same vendor for every site.
Standardization is where the savings live. One provider, one bill, one escalation path. When a store in one state goes down, the person who answers already knows the layout of a store in another. No re-explaining, no re-learning, no week-long installs.
The stores that get this right
The retailers with the fewest outages treat connectivity as a store-level requirement, not a corporate afterthought. They spec it into the opening checklist the way they spec the fire suppression and the alarm. They know what every store pays, what every store's backup is, and who owns the call when something drops.
If you can't answer those three questions for your fleet — what does each store pay, what's the backup, who do I call — then your connectivity is already costing you more than it should.
An hour of downtime is rarely just an hour. The register comes back, but the customer, the labor, and the inventory accuracy don't always follow.
If you want a network where a dead modem is a non-event instead of a fire drill, let's map out what that looks like for your stores.
Carter Dewey
Carter Dewey leads solution architecture at TrustedNetworx, helping multi-site organizations navigate telecom modernization, POTS replacement, and AI-powered operations. With deep experience across property management, senior living, hospitality, and healthcare, Carter translates complex infrastructure challenges into practical, phased migration roadmaps.