Build Profitable Growth From the Work You Can Deliver

Write Down Why the Margin Should Change

The model's warning is to fix service delivery before adding clients: more clients at 5% EBITDA is more 5%. That holds if the overall cost relationship stays the same. Growth can improve a margin if costs scale differently, but you need to show which costs change and why.

A larger revenue total is not a cost-reduction plan. If the forecast improves margin, identify the delivery work and overhead behind that improvement. Keep an untested assumption visible rather than entering a better percentage because the company will be bigger.

I'd first separate a deliberate growth investment from an unexplained operating shortfall. A company spending to acquire clients needs a return and funding plan. A desk consuming more work than the forecast allows needs its workload investigated before the next client is added to that forecast.

Find the Cost Problem in the Current Book

Gross margin is revenue less the direct costs of supplying the service, expressed as a share of revenue. For managed services, the model's target is 50% after burdened delivery labor and tools. That is different from EBITDA, which also accounts for sales and overhead.

Start with the client-level revenue and work records. Match the period and scope before comparing accounts. Have delivery check the hours and finance check that the relevant labor and tool costs are included once.

Ian Luckett recommends first examining which clients are profitable and which need improvement, then planning the conversations. That is a useful operating order. A client ranking should lead to an investigation of the cause, not an automatic price demand.

Check whether the agreement and actual work match. An unresolved scope issue needs a different correction from a recurring technical fault. Keep the action tied to the finding so you can later check whether the cost actually changed.

Use the Workload Sensitivity to Find What to Measure

The model assumes 0.5 tickets per seat per month for a well-run MSP; a typical MSP runs at 1.0 to 1.4. At 700 seats, the well-run assumption produces 350 tickets. At 1.0, the same seats produce 700.

That comparison holds seats constant. In the model's workload calculation, the desk moves from 4 technicians to 7. Its full US stable sensitivity puts the EBITDA reduction at $282,547, about 7.8 percentage points, compared with the well-run baseline.

The sensitivity shows why actual demand matters. It doesn't establish that reducing a ticket count will save that amount in your business. Check the work behind the tickets and whether staffing or other spending would actually change.

Don't subtract the sensitivity from a low-margin result that already includes the extra delivery cost. That would count the problem twice. Use it to identify the assumption worth measuring, then rebuild your own cost calculation from the records.

A lower count can also reflect a change in logging or work handled outside the desk. Keep the service outcome and recorded labor beside the count. You need evidence of reduced work, not a smaller number on a dashboard.

Choose the Repair Before Committing the Spend

I'd give each delivery improvement a named owner and an observable completion condition. If repeated faults are driving work, assign the technical correction and the follow-up check. If scope is the issue, assign the commercial discussion and have delivery cost the proposed position.

The model favors fixing a low-margin client through a useful refresh before repricing. Assess whether that correction addresses the measured cause and what it costs to deliver. A project label doesn't establish that future support work will fall.

Ray J. Green describes temporarily pausing sales while his business worked on onboarding and retooled its offer. That is his account, not a universal instruction to stop selling at a particular revenue level. Use your own delivery commitments to decide which start dates you can responsibly accept.

Keep the sales plan connected to that decision. The pipeline can continue while you arrange feasible delivery dates and resolve a constraint. Don't let the acquisition forecast assume capacity the delivery owner has not confirmed.

Price the Next Client as Additional Work

The next sale needs its own delivery plan. In the model, each new client takes 60 hours of onboarding by project engineers. Onboarding is quoted and waived, so the cost remains even though no onboarding fee comes in.

Include that work alongside paid projects. Don't put it into support-desk slack or count the same project-engineer time as available for both jobs. The person responsible for delivery should confirm how the added work fits.

Add the incremental sales and overhead costs to the plan without counting the same person twice. Then add expected revenue and its delivery cost on the dates you can support. An annual revenue total doesn't establish that cash arrives before the work must be funded.

Approve Growth From the Rebuilt Forecast

The model's 20% EBITDA target and 10% danger line give the owner a planning comparison after fair-market owner pay. The danger line is not break-even. A forecast above it still needs credible cash timing and delivery capacity.

Have finance show the baseline result and the specific changes behind the proposed growth case. Keep uncertain delivery savings separate from savings already demonstrated in the current book. If a planned improvement doesn't occur, the forecast should show what that does to the decision.

Start with the largest unexplained delivery difference in your current records. Assign its investigation before approving the spending that relies on it being fixed. Resume that approval discussion with the corrected workload and cost calculation in front of you.