What actually happens when your digital signage loses internet

Ask any digital signage vendor whether their product works offline and you will get a yes. Ask what happens when the router reboots and the answers diverge sharply, because "works offline" covers at least three different products.

The three meanings

Cached playback. The player downloaded the content earlier and keeps playing it. This is what almost every cloud product means. It works, and it is genuinely useful — a shop with flaky Wi-Fi will not see a blank screen. But the product still requires an account, a server and a connection to do anything other than continue.

Degraded management. The screen plays, but you cannot change it. Everything you might want to do during an outage — pull a promotion that is now wrong, put up a notice about the outage itself — needs the thing that is down.

Actually local. The content, the schedule and the management interface all live on the device. The internet is something individual widgets use, not something the product needs to stay alive.

Most of the category is the first. Some is the second. Very little is the third, which is odd, because the third is what a lot of buyers assume they are getting.

The questions that separate them

If you are evaluating, these four answers tell you which product you are looking at:

1. Can I change what is on the screen while the internet is down? If no, management is cloud-only.
2. Where is my layout stored? If the only copy is on their server, the screen is a viewer, not a player.
3. What happens if I stop paying? With a cloud product, usually nothing, immediately — the screen goes to a placeholder or stops updating. That is not a threat, it is just how a subscription works, but it is worth saying out loud before you deploy forty of them.
4. Does the device need an account? An account is a dependency with a support desk attached. Password resets are a real operational cost at scale.

What a screen should do when the feed dies

Whatever the architecture, one behaviour matters more than the rest: a widget whose data source has gone away should hold its last good value, not blank and not show an error.

In a public space, a board that is an hour out of date is almost always better than a board that is empty or showing a stack trace. A departure board with stale rows is a board somebody can still use with judgement. A blank one is a broken screen, and it makes the whole installation look unmaintained.

The corollary: if staleness would be dangerous — a queue number, a platform change — put a visible timestamp next to it so anyone reading knows how old it is.

The bit nobody puts on their pricing page

Connectivity has a cost that never appears in the per-screen comparison. A screen in a warehouse, a basement pharmacy, a marquee or a listed building may need a network run, a bridge, or a 4G router with its own SIM and its own monthly bill. For a single sign, that can exceed the signage subscription.

A product that does not need the connection at all removes that line entirely — which is often a bigger saving than the software.

Practical advice

Whatever you buy, do this before you deploy at scale:

• Put one screen up and unplug the network for a day. Watch what happens at hour one and hour twelve.
• Pull the power out of the wall while it is running, then plug it back in. If the board does not come back on its own, you have a site-visit problem, not a software preference.
• Check what the screen shows when a data feed 404s, not just when the whole network is gone. That is the failure that actually happens.

We built UniKiosk as the third kind, because the screens we kept being asked about were in places where the second kind quietly did not work. But the test above is worth running on whatever you choose.