Why the Week Is the Right Cadence
A month is too long. A client who's frustrated in the second week has usually made up their mind by the time a monthly report lands on your desk, and by then you're reacting to a decision instead of a signal. A day is too short. Daily numbers bounce around for reasons that don't mean anything, and watching them that closely just teaches you to ignore them.
The week is where a pattern first looks like a pattern. Two late starts on the same post, a report that got thinner three shifts running, a callout that took two hours to fill. None of those is a crisis on its own. Together, over seven days, they're the shape of an account that's about to go sideways, and the week is early enough to still turn it.
What Should You Actually Measure?
Keep the list short enough that you'll really look at it every week. The point isn't a dashboard with 40 boxes. It's a handful of measures that each answer a real question about whether the work got done.
- Coverage: how many posts went uncovered last week, how long each gap lasted, and who filled it. An open post is the failure a client is quickest to notice.
- Callouts: how many Officers called out, and how fast each shift got covered. The fill time matters more than the count.
- On-time starts: how many shifts started late or ran a no-show. A late start is a small number that predicts a big complaint.
- Reports filed: whether the required reports actually came in, and whether they show the Officer saw the site rather than just clocked the hour.
- Supervisor visits: how many site visits happened against how many were scheduled, and whether at least one was unannounced.
- Retention signals: who's inside their first 90 days, who's shown a warning sign, and who you'd struggle to replace this week.
- Renewals coming due: which Class D and Class G licenses, certifications, and insurance items expire in the next 60 days.
What the Numbers Are Actually Telling You
A weekly board is only useful if you read past the number to the decision behind it. Most bad numbers this week are really about a choice you made two or three weeks ago.
A climbing callout rate usually isn't a scheduling problem. It's a hiring or a morale problem that started upstream, and the schedule is just where it finally showed. Late starts rarely mean lazy Officers. They mean the Officer doesn't know the post well enough to respect its start time, which points back at the account startup. Reports that thin out don't mean the Officer stopped working. They mean nobody on your side was still reading them, so the standard quietly dropped. And a license you found out about after it lapsed isn't a compliance surprise. It's proof the tracking lived in someone's head instead of on the board.
Average Operator Versus a Strong One
The measures aren't the difference. Every operator can count bodies on posts. The difference is what each one does with the number, and whether they look before the client makes them look.
| What you're watching | Average operator | Strong operator |
|---|---|---|
| Coverage | Counts whether posts were staffed | Tracks how each gap got filled and how long it stayed open |
| Reports | Confirms a report exists | Reads whether the report shows the Officer saw the site |
| Supervision | Visits a site when there's a complaint | Visits on a schedule, including visits nobody expected |
| Retention | Notices when someone quits | Watches the first-90-day signals before they quit |
| Renewals | Learns a license lapsed when the state says so | Sees Class D and Class G expirations weeks out |
I check us against the same standard we set for everybody else, and I do it on purpose, every week. It's easy to measure your Officers and never turn that same look on the operation behind them.
If the reports thinned out, that's not just the Officer. That's whether anybody on my side was still reading them. If callouts climbed, that's whether we hired right and scheduled right weeks ago. The number on the board is almost always telling me something about a decision we made upstream. So I read it as a question about us first, then about the post.
Run Through Your Own Board
Before your next week starts, run through these and be honest about which ones you can answer without making a phone call:
- Do you know how many posts went uncovered last week, and how each gap got filled?
- Can you see which Officers started late, without asking a supervisor first?
- Did every scheduled supervisor visit actually happen, including at least one nobody expected?
- Do you know which licenses and certifications come due in the next 60 days?
- Would a rise in callouts show up on your board this week, or only after you'd already lost the account?
How We Handle It at ARDENT
We treat the weekly read as an operating discipline, not a report we file for someone else. Officers submit reports through the shift, so coverage and site activity are visible while the week is still running, not reconstructed after it. Supervisors work to a visit schedule that includes unannounced stops, because a visit the Officer expects only proves they can look good when they know you're coming.
Renewals live on the board, not in someone's memory. Class D and Class G licenses renew every two years, and we track them as an operational responsibility so a lapse never becomes the client's problem. And when a number moves the wrong way, we look upstream first, at the startup, the hire, or the schedule that produced it, because that's usually where the fix actually is. The goal is simple: catch the small thing on a Monday instead of hearing about the big one from the client.
Key Takeaways
- The right weekly measures predict a lost account before the client feels it. Coverage, callouts, on-time starts, reports, supervision, retention, and renewals cover most of it.
- Keep the list short enough that you'll actually look every week. A board you ignore is worse than no board.
- Read past the number to the decision behind it. This week's bad number usually started as a choice two or three weeks ago.
- Track renewals where you can see them, not where you have to remember them.
Frequently Asked Questions
How Long Should a Weekly Review Take?
Once the numbers live in one place, the read itself is short, often 15 to 20 minutes. Most of the work is setting it up so the measures pull automatically instead of being chased down each week. If your review takes hours, the problem is the plumbing, not the discipline.
Isn't This Just for Big Operations?
It's more important when you're small. A single owner covering a handful of accounts feels every uncovered post and every lost client directly, so an early signal is worth even more. The list scales down cleanly: fewer posts means fewer numbers to watch.
What If the Numbers All Look Fine but a Client Is Still Unhappy?
Then you're missing a measure, usually a client-side one. Add a weekly note on complaints, escalations, and the last real conversation with each account. Operational health and client satisfaction usually track together, but not always, and the gap is worth watching.