Before you talk to us

What your tools are not telling you

Not a sales claim — a list of specific things every MSP stack reports as fine, and how to check each one yourself before you talk to us.

THE SITUATIONMOST AGENTSTHIS ONE host does not answer a ping 0 ms unreachable disk will not report SMART ok unavailable throughput cannot be measured 0 Mbps no measurement a hop in a traceroute is silent dropped from the list silent nobody is logged on to run a command runs as SYSTEM says there is no user context A zero and a did-not-measure look identical on a dashboard, and only one of them is true. Every figure in the left column will be believed by somebody.

Check these yourself. Today. Without us.

Every item below is something your current tools almost certainly report as healthy. Each takes minutes to verify by hand — and if they all come back clean, you have an unusually well-run practice and you do not need us.

We would rather you found nothing than took our word for it. This page exists to be checked, not believed.

installed already installed reboot required failed most agents report this as "patched" The difference between the second and the third is a machine that is not protected and says it is. from the description on agents.deploy_patch
THE TOOL SAYS WHAT IS TRUE 47 of 47 compliant 6 behind a pending reboot 40 seats contracted 44 accounts live $312/mo USD 731 nights succeeded 0 restores ever attempted Every one of these is checkable. None of them is checked by the tool that reports it.

1. Pick five machines your patch tool says are compliant

Log into each one. Check the patch is actually installed and the machine has rebooted since.

What you will probably find: at least one where the patch is staged and pending reboot. Pending reboot is not protected — and most patch reports count it as compliant.

2. Restore one backup. Any backup.

Not a test-restore report. An actual restore, to somewhere, of something you can open.

The uncomfortable question: when did you last restore from each client's backup product? If the honest answer is at the last incident, then every green tick since is a vendor's opinion, not a verified state.

3. Count the seats on one client's agreement

Open the contract. Count what it says. Then count what is actually active in their tenant and on their network.

This is the one that usually finds money. Agreement drift runs 3–8% of contracted value across the industry, and it accumulates silently because nobody reconciles a signed number against a live one.

4. Ask who approved your last domain controller reboot

Not who did it — who authorised it, and where that authorisation is written down.

Most MSPs discover the answer is "we've always done it that way", which is a fine answer until a client asks it during an incident.

5. Look at what your RMM says is monitored

Then check the agent is actually reporting. An asset that stopped checking in six weeks ago often still appears in the inventory as managed.

Absence of alerts is not evidence of health. A machine that went quiet looks identical to a machine with nothing to report.

6. Find one thing you are paying a distributor for and not billing

Compare one month's distributor invoice against what you billed. Line by line, for one client.

This is tedious, which is exactly why nobody does it, which is exactly why it accumulates.

What we do about it

All six run continuously, on every client, and each finding names the assets behind it and what it costs.

Not because it is clever. Because nothing else checks.

See it on your own estate

Thirty days, read-only. Nothing changes and no client is contacted.

Or let Genie check all six for you

Thirty days, read-only, on your real estate. Nothing is changed.