The charter
One page both of us can be held to.
Five ranked priorities. The criteria each one is judged on. Fifteen dated deliverables across four stations. Twelve metrics with a name against each. Nothing in it is a description of effort.
How it is written
A criterion you cannot fail is not a criterion.
Every line has to survive one test: could two people who disagree about everything else still agree on whether it was met?
It names a state, not an effort
“We will improve onboarding” is effort, and it is met the moment anyone tries. “A new client is live in five working days with no escalation” is a state. Only one of the two can be failed.
It has one owner
A person, not a department. They knew before it was written down, and they should never learn at the review that a number belongs to them.
It has a baseline
Taken before anything changes, from your own systems. A criterion with no number behind it is an argument you have scheduled for nine months’ time.
It is worth the trough
Every criterion costs before it pays. If nobody can say what the return is on the far side, we cut it rather than carry it for six months.
Twelve to twenty criteria across the five priorities. If it needs more than that, the priorities are wrong, not the writing.
The five
Ranked, because they will conflict.
Protecting what works and making the desk cheaper pull against each other in week eleven. Ranking them in week three means that conflict has an answer before anybody is emotional about it. Yours will be worded for your business; the shape below is where most MSPs land.
Protect what already works
Growth is the most likely thing to break the thing that got you here. The service that answers on the second ring, the technician who already knows the site, the client who never has to explain twice.
Judged on- The service your best clients buy is unchanged
- New hires are held to the bar the first ten were
- Accountability rises without another layer of management
- Anything you acquire adopts your standard, not the reverse
Put a name on every number
The largest gap in most MSPs is not capability, it is ownership. Nobody is idle. It is just that no single person can be asked why a number moved.
Judged on- Every metric has one owner, not a committee
- Expectations are written down and measurable
- Managers run to a scorecard, not to memory
- The weekly review names people, not departments
Remove the single points of failure
The riskiest thing in the business is usually a person rather than a system. Most of it is invisible until they take annual leave.
Judged on- No process depends on one individual’s recall
- Client knowledge sits with a team, not a technician
- Every internal audit has an owner who runs it
- You can be unreachable for two weeks
Give the client something to sign
Clients cannot accept risk they have never been shown, and you cannot defend a recommendation nobody recorded them declining.
Judged on- Every managed client has a current posture review
- Recommendations are standard, not improvised per account
- Approved, deferred and declined are all recorded
- The record survives the account manager leaving
Make the desk cheaper to run
Service delivery is the engine and it is where the margin leaks. Usually into work that should never have reached a person at all.
Judged on- Less noise reaches a human at all
- Time entry is complete without anyone chasing it
- More of the week is proactive work
- Load is visible before it becomes an escalation
The 180 days
Four stations, fifteen deliverables.
A station is a date with artefacts attached. Each one either exists on the day or it does not, and the review says which. There is no percentage complete anywhere in this engagement.
Assess
Understand who runs what, and what happens when they stop.
- Deliverable: Manager capability assessment
- Deliverable: Operational baseline scorecard
- Deliverable: Owner dependency map
Diagnose
Score the operation and rank the gaps by what they cost you.
- Deliverable: Operational maturity assessment
- Deliverable: Ranked gap list
- Deliverable: Staffing and capacity read
- Deliverable: Alerting and noise review
Design
Build the future state and the accountability that holds it up.
- Deliverable: Future-state structure
- Deliverable: Accountability framework
- Deliverable: Build queue, sequenced
- Deliverable: Improvement roadmap
Execute
Institutionalise it, then hand back the chair.
- Deliverable: Operational playbook
- Deliverable: Scorecard your leads run
- Deliverable: Scalability roadmap
- Deliverable: Next-year operating plan
Why the dip lands between Day 60 and Day 120
Standardising a process is slower than the way people were already doing it. That is not a failure of the design, it is the cost being paid. Ticket counts rise, technicians complain, and someone suggests going back.
We say the month it will happen, in the charter, before it happens. Naming it in advance is the difference between a planned cost and a crisis.
What the stations are not
They are not billing milestones and they are not gates you have to pass to keep working with us. Nothing stops at Day 60 because a deliverable slipped.
They are the dates on which we both look at the same list and say out loud what exists. Slippage is normal. Slippage nobody named is the problem.
The standing questions
Five questions every charter has to answer.
We ask the same five in every engagement. Your answers are the diagnostic; ours are below, and they are what we expect to find before we have met you.
What actually limits taking on the next twenty clients?
Almost never headcount. Usually three people are the process, escalation is a habit rather than a rule, and nobody can say which step in a ticket’s life is the slow one. We answer this from your data, not from the leadership team’s impression of it.
Where is accountability strongest, and why does it stop there?
In most MSPs, finance and admin are tight and service delivery is loose — because finance has a number that closes every month and service has a feeling. We work out what makes the tight part tight, then copy it, rather than importing somebody else’s framework wholesale.
Which processes exist only in one person’s head?
The dependency map is uncomfortable to look at and it is the most valuable page in the diagnostic. We name individuals, list what leaves with them, and rank it by what the fortnight after they resign would cost you.
Are you using what you already pay for?
PSA and RMM usually run at a fraction of their capability, with the knowledge held by the two people who configured them years ago. Before anyone buys anything new, we establish what the current stack could do if somebody owned it.
Which decisions still need you personally?
Every owner-dependent decision is both a ceiling on the business and a tax on your week. We list them and sort them three ways: delegate now, delegate once a control exists, keep. Your largest client relationships are the last to move, not the first.
And the one we cannot answer for you
What are you willing to be worse at for a quarter? Every charter trades something. If the answer is nothing, the honest recommendation is to keep running the business the way you are and spend the money elsewhere.
The rhythm
Twenty-five minutes a week, against the same page.
A charter read twice a year is a document. Read weekly, it is a control.
- Same time, same agenda, twenty-five minutes. It does not expand to fill the hour.
- Scorecard first. Twelve numbers, and owners read their own out loud.
- Off-track is not a discussion. It is an owner, a cause and a date.
- Anything that cannot close in the room becomes a deliverable against a station.
- We chair it until your leads chair it better than we do, and then we stop attending.
Exit criteria are agreed in week five. If we are still in your stand-up in month eighteen, something has gone wrong.
- Shipped: Future-state structure
- Shipped: Accountability framework
- Shipped: Build queue, sequenced
- Shipped: Improvement roadmap
- Not shipped: Acquisition integration plan — slipped, owner named, new date agreed
- Not shipped: Centralised services scope — blocked on a hiring decision
Illustrative. Two open items with a name and a date on them is a healthy review. Six out of six is usually a charter that was written too easily.
Scope
What never goes in the charter.
Worth reading before you book. It saves both of us a call.
- Stack replacement
- We will not tell you to rip out your PSA or RMM. Migrations create eighteen months of trough for a benefit that is usually available without them.
- Resold tooling
- We take no vendor commission and resell nothing. If a product is the right answer we will say so and step out of the transaction.
- Staff augmentation
- We are not a NOC, a SOC, or overflow engineers. We change how your operation runs; we do not run it for you.
- Documentation for its own sake
- We write down what has to be written down and no more. A 200-page runbook nobody opens is bureaucracy with better formatting.
- Criteria that depend on us
- Nothing in the charter can be met only while we are in the room. If a criterion needs our attendance to stay green, it is written wrong.
- Open-ended retainers
- The charter has an end date and so do we. Renewal is a new charter with new priorities, not a continuation of the old one.
Start small
The charter is the first deliverable.
Three weeks, fixed fee. You end it holding the five priorities, the criteria, the OML baseline and the dates — and you are free to run the rest yourself.