"When the power goes out: what in your store keeps working and what stops"

Storms, grid failures, a substation fire, a scheduled outage — the cause varies by country, but the shape of the problem does not. And unlike most seasonal peaks, this one does not start on a date: demand for power banks, battery stations, thermoses and heaters jumps with the news rather than with the calendar. Which is exactly why you cannot prepare at the last moment: by the time demand appears, the stock is either on your shelf or it is not.

But there is a second half to this season that gets thought about less. Not "what to sell" but "will I be able to ship it". And for that it helps to break the store down link by link, because the links do not all stop, and they do not stop at the same time.

What does not lose power

The site itself. It does not sit with you but with your host or platform, in a data centre with its own power. The buyer still sees the catalog, still adds to the cart and still places the order — even while your lights are off.

That matters, because the season's main mistake follows from it: orders keep arriving and you cannot see them. The pause is not in sales but in processing — and the buyer has no way of knowing.

What stops

• You. The computer orders are processed on, the router, sometimes mobile signal along with the nearby masts.
• Printing labels and documents. A printer with no power is a parcel that did not ship, even when everything else was ready.
• The card terminal for click-and-collect, and any till tied to a wired connection.
• Receipt generation, if your receipt software runs on that same computer with no backup channel.
• The warehouse, if it is dark enough that picking an order is physically impossible.

Online card payment on the site, incidentally, does not stop — it happens between the buyer, the payment service and the bank, and needs none of your electricity. A store that only takes cash on delivery or a manual bank transfer loses more in this season: a transfer requires somebody to look at it and confirm.

The minimum that closes most of the problem

1. A laptop instead of a desktop for processing orders. The cheapest and fastest change on this list.
2. A UPS on the router. A router draws very little, so a small unit lasts hours rather than minutes.
3. Mobile data as a second channel and a tested switch to it. Tested means you have actually tried it, not that you think it will work.
4. A charged phone with access to the admin panel or CRM. Confirming an order and replying to a buyer can be done from a phone; picking a parcel cannot.
5. Online payment on the site, so that getting paid does not depend on someone being awake to look at a transfer.

The page almost nobody has

A small separate page about how the store operates during outages: that orders are accepted around the clock, what hours you call back, how long shipping takes right now, and what to do if there is no reply for longer.

It does three things at once. It removes part of the question load from your inbox. It calms a buyer who can see they have not had a reply for an hour. And, as a side effect, it is one of those texts people genuinely search for — unlike an "About us" page.

While you are there, check whether the site even has the basic shipping, payment and returns pages. Often it does not, and in a difficult season that shows up sharpest.

An auto-reply that is not a box-tick

The automatic reply on email and messaging should carry something specific rather than "thank you for contacting us". A useful text carries a deadline: "I can see your message, I will reply within two hours, and your order can be placed without waiting for me". A deadline you actually hold to works better than any apology.

What not to do in this season

Migrate the store to another platform, restructure the catalog, or roll out a CRM. Demand jumps without warning, and any large change lands precisely in the week when it is most expensive. That is summer work — the lowest-traffic time of the year and the best time for changes.

Where to start

Walk the five links above and mark which ones you have actually tested, rather than which ones "should work". Switching to mobile data and printing labels are the two places where it most often turns out nothing has been tested at all.

Then write that opening-hours page. Half an hour, and it will serve you for more than one season.