1,000 views
Almost every number on a pet business dashboard is something you want to improve. Overdue visits is the exception. It is the one metric with a correct value, and that value is zero. Any figure above zero means a scheduled visit passed its window without being confirmed as completed, and in a business where the service is looking after a living animal that has been left alone, that is not a reporting inconvenience. It is the single highest severity signal you can have.
The reason this KPI earns permanent space at the top of the dashboard is not that it goes wrong often. In a well run operation it almost never does. It is that when it does go wrong, minutes matter, and no other metric will tell you in time.
What this KPI actually measures
Overdue visits counts scheduled services whose time window has fully elapsed without a completion record. It is a real time operational count, not a historical average, and it should be visible continuously rather than reviewed monthly.
Note carefully what it is measuring: the absence of a confirmation, not the absence of care. In practice, a large share of overdue visits are visits that happened perfectly well and were simply never marked complete, usually because a sitter had no signal or forgot to close the visit in the app. That distinction matters for how you respond, but it does not reduce the urgency, because from where you are sitting the two cases are indistinguishable until you check.
How to calculate it
Define a completion window for every visit type. A midday dog walk booked for 12:00 to 14:00 is overdue at 14:01 if no completion has been logged. An overnight stay needs a defined check in confirmation time rather than being treated as a single twelve hour block, otherwise a problem at 9pm will not surface until morning.
Then count every scheduled visit past its window without a completion event. Add a small grace period if you must, but keep it short. Fifteen minutes is reasonable; an hour is not, because an hour of grace is an hour you have chosen not to know about a pet that may not have been fed.
Track two versions. The live count is what should sit on the screen, and it should be zero. The historical count, aggregated weekly, is what tells you whether your process is degrading, and it is the version worth reviewing in a monthly operations check.
A worked example
A team of four covers 92 visits on a Wednesday. At 14:30 the dashboard shows two overdue. The first is a 12:00 to 14:00 walk assigned to a sitter who is known to lose signal in a particular neighbourhood; a message confirms the walk happened at 12:40 and the record is closed manually. The second is a 13:00 cat feed with no response to two messages.
That second one is escalated: the owner calls, the sitter does not pick up, and a second sitter is dispatched to the property. It turns out the first sitter had a flat tyre and had assumed she would get there before anyone noticed. The cat is fed at 15:10, roughly two hours late, and the client is told before they think to ask.
Over the week the historical count shows 11 overdue records against 480 visits, of which 9 were logging failures and 2 were genuine delays. The logging failures are not harmless: they are noise that will eventually cause a real overdue visit to be ignored, which is exactly how serious incidents happen in this industry.
What a healthy number looks like
Live, the healthy number is zero, at every moment of every day. There is no acceptable standing level.
Historically, a well run operation should see overdue records on well under 2 percent of visits, and genuine late or missed visits on under 0.2 percent. If you are above 5 percent on the logging side, the problem is your process or your software rather than your sitters, and it needs fixing as a system rather than being handled as a series of individual conversations.
The most useful secondary measure is time to resolution: how long from a visit going overdue to it being either confirmed or corrected. Under fifteen minutes is a competent operation. Over an hour means the alert is not actually reaching a human who can act on it.
How this looks different by business model
Solo operators often assume this metric is irrelevant because they know where they are. That is true for the visits themselves and false for the client experience, since the completion record is usually what triggers the client update. A solo sitter with a high overdue count is not neglecting pets, but is very likely eroding trust by leaving owners without confirmation.
Team based businesses are where the metric becomes critical. Once you are not personally performing the visits, this count is your only real time evidence that the service is being delivered, and it should drive an active alerting workflow rather than a passive number on a page.
Overnight and house sitting carries the highest stakes, because the gap between visits is longest and the pet is genuinely alone in between. Those bookings deserve tighter windows and their own escalation path. Businesses running high volumes of same day work also see elevated overdue rates simply because rushed scheduling leaves less slack in the day.
What actually moves this number
Four things cause overdue visits, and they need different fixes. Scheduling that is too tight is the most common: routes with no travel buffer guarantee that one delay cascades through an entire afternoon. Compare your scheduled load against visits today and your realistic capacity before blaming anyone.
Second is app and connectivity friction. If closing a visit takes six taps and fails on poor signal, sitters will defer it and then forget. Third is unclear expectations; if nobody has ever said that closing the visit is part of the job rather than paperwork, it will be treated as paperwork. Fourth is genuine incident, which is rare and is precisely what the whole system exists to catch.
Reading this number next to visits completed today makes the difference between a logging problem and a delivery problem obvious within seconds, because a logging failure leaves the two counts inconsistent while a delivery failure moves them together.
Common mistakes when tracking this KPI
The first mistake is reviewing it retrospectively. An overdue visit discovered on Friday for a Tuesday booking has no value except as an incident report. This is a live metric or it is nothing.
The second is generous grace periods, which are simply a decision to be told about problems later. The third is treating logging failures as harmless; every false alert trains you to ignore the real one.
The fourth is having no escalation path. A number on a screen that nobody is required to act on is decoration. There should be a written sequence: message the sitter, call the sitter, contact the backup, attend personally, inform the client.
The fifth is not telling the client. Owners forgive a late visit far more readily than they forgive finding out about it themselves, and the businesses that handle this well often convert an incident into a stronger relationship than they had before.
How it connects to the rest of your dashboard
Overdue visits is the leading edge of everything reputational. It feeds directly into retention, review scores, and rebooking, because a missed or unconfirmed visit is one of the very few things that will end a pet care relationship in a single event. Businesses with persistent overdue counts see it surface months later as unexplained churn.
It also correlates with average response time, since both measure the same underlying capability: whether the operation notices things quickly and acts on them. An organisation that is slow to answer enquiries is usually slow to notice a visit that did not happen.
Turning this into a repeatable habit
Put the live count somewhere it cannot be avoided, and set an automatic alert at the moment a visit goes overdue rather than relying on anyone remembering to look. Assign an owner for that alert during every operating hour, including weekends and holidays, when both the volume and the risk are highest.
Then hold a five minute weekly review of the historical count. Look at which sitters, which routes, and which times of day produce the records. The pattern is usually concentrated rather than random, and fixing one route or one device often removes most of the problem permanently.
Action checklist
Define a completion window for every service type and keep the grace period under fifteen minutes. Get the live count onto a screen you actually see. Write the escalation sequence down and make sure every sitter knows it. Separate logging failures from genuine delays in your weekly review. Set a target of zero live and under 2 percent historically, and treat any breach as an operational defect rather than an individual failing.
The bottom line
Overdue visits is the one number on your dashboard with a right answer. Most of the time it will be zero and you will wonder why you bother watching it. You bother because pet care is a trust business, because the animal cannot tell anyone it was not fed, and because the difference between a minor delay handled well and a serious incident is almost entirely a question of how quickly somebody noticed.
