What is recovery time objective?

Setting it honestly means costing an hour of downtime, which most businesses have never actually done.

The answer differs sharply by system. An hour without your accounting software is a nuisance; an hour without your phone during a busy morning is lost customers.

Targets are often written for the systems IT owns and never for the telephone, which is the one customers experience directly.

Achieving a short target costs money in advance. Achieving a long one costs money on the day, and only one of those is budgeted for.

Say your dental office takes about 18 calls an hour between 8 and 10 am. Roughly a third are people booking treatment, so call it six. If half of those give up and try the next practice, you lose three bookings an hour. At a first-visit value of $250, that's $750 per hour of silence. Your practice software being down for the same hour costs far less, because the front desk can write on paper and catch up later.

Those two figures give you two targets. The phone line gets 15 minutes, since you'll be close to $200 down by then. Your practice software gets four hours. Now your plan has an order: whoever finds an outage forwards the main number first and calls the software vendor second. Plenty of written plans have this backward, because the software came with a support contract and the phone didn't.

A target is a wish until you've timed it. Pick a quiet Tuesday, pretend the main line is dead, start a stopwatch, and have the person who'd really be there do the fix. Count everything. That includes six minutes spent finding the carrier login, and two more waiting for a code texted to an owner who's on a plane. First runs of a "15-minute" fix often take 40.

In a real incident the clock starts when the failure happens, which can be long before you notice. If your line dies at 7:50 and the first person arrives at 8:30, you've burned 40 minutes before anybody lifts a finger. Detection belongs inside the target, so a daily test call or a carrier alert is part of meeting it.

Short targets aren't worth buying for everything. If a day without your reporting dashboard costs you nothing but mild annoyance, write "one week" beside it and spend the money elsewhere.

GreetKeeper makes no recovery time commitment and offers no service level agreement, which is the honest position for a product this new.

Your practical protection is a carrier-level fallback, because it runs on the network rather than inside anybody's system.

Point it at a mobile you will actually answer, and check once a year that the rule is still there.