Clients rarely leave over one bad month. They leave after six months of small signals that each sat in a different system, and nobody put them side by side.
Most MSPs find out a client is unhappy when the notice letter arrives. Looking back, the signs were there: tickets per user creeping up, an invoice sixty days overdue, a new finance director nobody had met. Each signal lived in a different system, and no one person looked at all of them. This free monthly client health check checklist gives account managers and service leads an internal routine for doing exactly that, one client at a time, before anything goes to the client. It reviews ticket trends and SLA status, estate health from the RMM, security signals, backup success, licence and asset changes, and the relationship and commercial picture. Each area gets a red, amber or green rating and the client gets an overall score. When the score is red, a conditional phase builds a recovery plan, gets it approved by the service manager and puts an executive call in the diary.
The health check is that question, asked every month. It is internal and candid: it records the noisy user, the overdue invoice and the champion who has just left, things that never belong in a client report. It reuses the outputs of those routines and produces two things: a score the service manager can compare across clients, and a short list of actions. Once a quarter, three health checks become the internal briefing for the MSP QBR Checklist, so the account manager walks into the meeting already knowing where the relationship stands.
Monthly service report
Written for the client
Audience: the client’s contact and decision-maker.
A health score only works if two account managers looking at the same client would give the same answer. Write the rules down, keep them short and apply them every month. The thresholds below are examples of common practice to start from, not a standard. Tune them after a few months of scores.
Area
Green
Amber
Red
Service desk
Tickets per user near the client’s own average, SLAs met
Volume up noticeably, a repeat issue open, a minor breach
A P1 breach, or volume rising for three months in a row
Estate
Agents reporting, patching inside your target
A handful of devices offline or failing patches
Unsupported operating systems or servers unpatched past your window
Security
MFA for everyone, detections resolved, Secure Score steady
Gaps in MFA, or a fall in Secure Score you can explain
An admin without MFA, an unresolved detection or a confirmed incident
Backup
Jobs succeeding, restore tested recently
Isolated failures, fixed within a day
A system outside its recovery point, or no recent restore test
Relationship
Good CSAT, contacts stable
A complaint or a lukewarm QBR
An escalation to your owner, or your champion has left
Commercial
Invoices paid on terms, scope matches the agreement
Late payment, or regular unbilled work
Invoices well overdue, or a request to cut scope or price
Do not average the six ratings. A client that is green everywhere except security, with a global admin account that has no MFA, is not “mostly green”. A common rule is that any red area makes the client amber at least, and a red in relationship or commercial, or two reds anywhere, makes the client red.
Signals that predict a lost client better than ticket counts
Your champion leaves. The person who chose you is replaced by someone who didn’t, and who may have used another provider before.
Payment slows down. A client who always paid in thirty days now pays in sixty, without saying why.
Requests for passwords and documentation. A sudden interest in admin credentials or network diagrams can mean another provider is being briefed.
QBRs get delegated or cancelled. The decision-maker stops coming, which usually means they have stopped deciding in your favour.
Reading the Microsoft signals correctly
Microsoft describes Secure Score as a measure of how many of its recommended actions have been taken, not as a measure of how likely a breach is, so record the trend and the reason for each change rather than chasing the number. Risky sign-in detail depends on licensing: with Microsoft Entra ID Free or P1 the risky sign-ins report shows limited information without risk levels, and full detail needs Entra ID P2. For operating-system risk, Windows 10 reached end of support on 14 October 2025. Organisations can buy Extended Security Updates a year at a time for up to three years, and the price doubles each year, so any Windows 10 device that is not enrolled belongs in the red column.
Why Run Client Health Checks in CheckFlow?
1
Every client checked, every month
A recurring schedule creates a health check for each client at the start of the month, assigned to the account manager with the data pull assigned to the service lead. A client nobody has looked at shows up as overdue, not as a surprise at renewal.
2
Red clients get a plan, not a worry
Conditional logic adds the recovery phase only when the score is Red, and the plan waits for the service manager’s approval before anything is promised to the client.
3
Scores you can compare across the book
The grid view lines up every client’s health check for the month, so the service manager can see which clients are red, which actions are overdue and which account managers are behind. Three months of scores make the QBR briefing write itself.
Monthly client reviews are one of the recurring processes the MSP process management guide recommends standardising. CheckFlow for MSPs runs them alongside onboarding, reporting, QBRs and renewals across every client.
A red score close to the end of a term should start the renewal early. The Managed Services Agreement Renewal Checklist picks up the health scores, and its retention phase takes over when a client has given notice.
It is an internal monthly review of one managed services client, done by the account manager or service lead. It brings together signals that normally sit in separate systems: ticket trends and SLA status from the PSA, estate health from the RMM, security and backup results, licence and asset changes, and relationship and payment signals. The output is a red, amber or green score and a short list of actions, with a recovery plan for any client scored red.
Should the client see their health score?
+
Usually not as a score. The health check records things that are unhelpful to share bluntly, such as a noisy user or slow payment, and a client who sees “amber” will spend the meeting arguing about the label. Share what the score is based on instead: the service, security and backup findings go into the monthly report, and the actions you agree become recommendations the client can see.
How long should a monthly health check take?
+
For a small or mid-sized client, often between fifteen and forty-five minutes once the data is easy to pull. Most of the time goes on Phases 2 to 4, and much of that is reading outputs that already exist, such as the SLA review and the daily backup log. If it takes much longer, the data pull is the problem: build saved reports in the PSA and RMM so Phase 1 is a few exports, not an afternoon.
Is a health score the same as CSAT or NPS?
+
No. CSAT and NPS tell you how users or the client contact feel about recent service, and they are one input to the relationship area. A health score combines that with signals the client never reports, such as estate risk, unresolved security issues, overdue invoices and the loss of your main contact. A client can give excellent CSAT scores right up to the month the new managing director moves to another provider.
What should we do with an amber client?
+
Treat amber as a warning with a deadline. Agree one or two actions in Phase 6 that address the reason for the rating, give each an owner and a date, and check them at the start of next month’s health check. A client that stays amber for three months without improvement is heading for red, and many MSPs make that a rule: three consecutive ambers count as red.
Is CheckFlow free for this template?
+
14-day free trial, no card required. The Business plan is $10 per user per month after the trial. Full details at checkflow.io/pricing.
Know Which Clients Are at Risk Before They Tell You
Free trial — no credit card required.
Do you like cookies? 🍪 We use cookies to ensure you get the best experience on our website. Learn more