Writing an MSP Business Plan

Why a lean operating plan beats a 40-page document

Nobody is going to read your 40-page business plan – not a bank, not a partner, and within three months, not even you. Business-school plan templates are written for raising capital; an MSP is a cash-flow business you bootstrap. What you need is a lean operating plan: a short document (a few pages, or a spreadsheet plus one page of narrative) that states who you serve, what you sell, what it costs to deliver, and what has to be true for the numbers to work.

The point of the plan is not prediction – it's falsifiability. Every section should contain assumptions specific enough that reality can prove them wrong, so you can correct course quarterly instead of discovering in month 18 that the model never worked. This matters more than it sounds: as of the most recent Service Leadership index (Q4 2024), 18% of MSPs ran at a loss. Profitability is not automatic in this industry, and the difference is usually a model someone actually checks.

The six sections that matter

1. Target market and niche. Define who you serve tightly enough to build a list: industry, company size, geography. "SMBs in my metro" is not a target market; "law firms with 10–50 seats within 25 miles" is. You don't need to commit to a vertical on day one – the common advice is start generalist-local, notice which vertical accumulates, then lean in – but the plan should name your starting focus and the criteria for narrowing. Vertical specialization is a real economic lever: specialized MSPs report up to 30% higher margins and 20–40% premium rates, with regulated-vertical contracts running $200–$400+/user/mo versus the $100–$250 generalist norm. See choosing a niche.

2. Service catalog and pricing. List exactly what's in your managed offering, what's out of scope, and what each tier costs. Fixed-fee per-user pricing ($100–$250/user/mo is the typical national norm as of 2026; $70–$150 at the low end for basic stacks) with a defined catalog is the standard – hourly billing caps your income at your hours and attracts clients who fight every ticket. Work through the service catalog and pricing models before setting numbers.

3. Financial model: MRR and gross margin per service line. Build the model on MRR, not projected total revenue. Recurring revenue is the business – it enables forecasting, hiring, and eventually valuation (managed-services recurring revenue is valued at roughly 4–6x annualized, project revenue at 0.5–1x). For each service line, model gross margin: price minus the direct cost to deliver (tool licenses per seat, labor hours, third-party services). Project work deserves its own line with honest margins – project-services gross margin collapsed from 23.0% to 12.9% year-over-year in the latest index, a strong argument against a project-heavy model.

4. Tool stack and cost basis. Your stack is your cost of goods sold, so the plan needs it itemized: RMM/PSA, EDR, backup, email security, documentation – typically $300–$800/mo per technician as of 2026. Standardize on one stack and treat the service like a product; a messy pile of one-off tools kills the margins you just modeled. Details in the MSP tool stack.

5. Sales pipeline assumptions. Write down where clients actually come from and at what rate: how many warm-network introductions you'll ask for, the size of your hyper-local target list (50–100 businesses of 5–25 employees is a common starting point), expected conversion. Referrals convert 3–5x better than cold outreach, so weight the plan accordingly. Be pessimistic: 1 in 3 MSPs name new-customer acquisition their single biggest challenge (Kaseya 2025 benchmark). Tactics in getting your first clients.

6. Hiring triggers tied to revenue. Don't plan hires by date – plan them by threshold. Common triggers for the first technician: roughly $100k–$125k in labor revenue, or 20–30 hours/week of steady overflow you're personally absorbing. Sanity-check against industry average revenue per employee of ~$142k – below that, each hire dilutes profit. A useful framework: the first hire should reach revenue covering their cost, then ~2x their cost, before you plan the next.

Modeling break-even honestly

This is where most first plans quietly lie to their authors. Three realities to build in:

• Client-level break-even takes 7–12 months. Onboarding a managed client typically consumes $10,000–$15,000 in labor, cleanup, and tooling before the relationship turns profitable. Model each new client as a cash outflow for its first several months.
• Growth consumes cash. Because of the above, a quarter in which you sign three clients is a quarter your bank balance drops. Your plan should show the trough, not just the trendline.
• Costs front-run revenue. Vendor commitments, insurance, and legal setup all hit before the first invoice clears. Bill monthly in advance and keep 3+ months of operating expenses in the business account – put both in the plan as policy, not aspiration.

Then compute two numbers: company break-even MRR (fixed monthly costs plus your minimum salary, divided by blended gross margin) and the month you expect to cross it given your pipeline assumptions. If the answer requires conversion rates you've never achieved, the plan fails its own test – fix the pricing, the costs, or the runway, not the spreadsheet formatting.

Sanity-check against real benchmarks

Your model's outputs should be plausible against what the industry actually achieves. Key figures from the Service Leadership / ConnectWise index (Q4 2024) and related benchmarks:

If your plan shows 65% gross margin on managed services or 30% EBITDA in year one, you've mis-modeled your delivery costs – average shops run ~46% GM, and the best sustained operators hold 50%+ GM and 19%+ EBITDA. Conversely, if your modeled GM is below 40%, your pricing or stack costs need work before you sign anyone. Also note the market context: managed-services revenue growth has slowed to ~1% worldwide (0.2% North America), so growth comes from taking share and adding security services – 67% of MSPs list security among their fastest-growing lines. A plan assuming a rising tide is a plan assuming wrong. Track your own numbers against MSP KPIs and benchmarks once you're operating.

Revisit the plan quarterly

A lean plan only beats the 40-page version if you actually re-open it. Put a quarterly review on the calendar – the same discipline you'll later apply to client QBRs, applied to your own business:

1. Compare actuals to assumptions: MRR, gross margin per line, pipeline conversion, churn.
2. Kill or correct the assumptions reality disproved – reprice, cut a tool, narrow the niche.
3. Check hiring triggers against current labor revenue and overflow hours.
4. Re-run break-even with the corrected numbers.

One page of notes per quarter is enough. The plan that gets revised badly every quarter beats the perfect plan nobody touches.

Where to start

Open a spreadsheet, not a word processor. Model your first ten clients at your intended price point, subtract your real stack and labor costs, and find your break-even MRR and the month you cross it – hedging every assumption you haven't tested. Then write one page naming your target market, catalog, pipeline plan, and hiring triggers. If the numbers survive contact with the benchmarks above, you have a plan; if not, better to learn it now. From there, startup costs will pressure-test your budget and how to start an MSP lays out the execution sequence.