A dollar of contracted monthly recurring revenue and a dollar of project revenue look identical on your P&L and sell for very different prices. If you take one idea from this entire blog into a negotiation, take this one, because revenue mix moves MSP valuations more than any other single factor, and it’s also one of the few factors you can deliberately change with enough runway.
Here’s the reasoning buyers apply, and the math underneath it.
The question every buyer is really asking
A buyer pays today for earnings that arrive tomorrow, so every valuation conversation reduces to one question: how confident am I that this revenue shows up next year without anyone re-earning it?
Contracted managed services revenue answers that question well. The client signed an agreement, the service renews by default, and next January’s invoice goes out unless something actively breaks the relationship. The revenue is there until someone makes a decision to leave.
Project revenue answers it badly. The migration you billed in March ended in March. For that dollar to exist next year, someone has to find another project, scope it, win it, and deliver it. The revenue is absent until someone makes it exist. Time-and-materials work sits in the same bucket with extra volatility, since it depends on things breaking at a billable rate.
Same dollar this year. Completely different probability of being a dollar next year. Buyers price probability.
What that does to the multiple
The market expresses this as different effective multiples on different revenue streams. In practice, when buyers value an MSP they weight contracted recurring earnings at the full multiple, weight project and T&M earnings at a substantial discount (half or less is a common shorthand), and weigh hardware resale barely at all.
Watch what that does to two identical-looking businesses. Both earn $600k of adjusted EBITDA on $2.5M of revenue:
- Shop A is 75% contracted MRR, 25% projects. Most of its earnings are the durable kind, and a buyer can comfortably pay a full multiple, call it 5x, on the bulk of them.
- Shop B is 30% contracted MRR, 70% projects and T&M. Most of its earnings have to be re-won annually, so buyers either discount the multiple hard or structure the price so the seller carries the re-winning risk through an earnout.
Run any reasonable weighting and Shop A comes out worth 30% to 50% more than Shop B, on identical earnings. It also attracts more buyers, closes with more cash up front and less earnout, and survives diligence with fewer surprises, because contracted revenue is verifiable in a way pipeline optimism never is. The two shops report the same adjusted EBITDA, and the market treats them as different species.
What counts as recurring (a stricter definition than you’d like)
Sellers use “recurring” generously. Buyers don’t, and diligence applies the strict version, so you should too. In descending order of credit:
Fully credited: contracted managed services with a term and auto-renewal, per-seat or per-device, where the agreement can be assigned to a buyer. Contracted security services, backup, and compliance monitoring sit here too.
Partially credited: month-to-month managed clients with long tenure. The history helps, but a client who can leave with thirty days’ notice is priced accordingly. Recurring-ish license resale (M365, for instance) gets counted but at thin credit, because the margin is small and the relationship is really Microsoft’s.
Not credited: “they call us every year” project regulars, warranty renewals you broker, and the anchor client who has always done a big refresh. Real money, real relationships, and none of it contractual, so none of it counts as recurring no matter how reliable it has felt from your chair.
The honest self-audit takes an afternoon: recompute your MRR percentage using only the fully-credited definition. Most owners find the strict number is 10 to 20 points below the number they’ve been quoting, and it’s better to meet that fact now than across a negotiating table.
What a healthy mix looks like
Benchmarks vary by market and service line, but the working bands we see: contracted recurring revenue above 70% of total reads as a managed services business and gets priced like one. Between 50% and 70% is normal and unremarkable. Below 50%, buyers start reading the company as a project shop with some contracts attached, and the valuation conversation changes character.
Projects aren’t bad, to be clear. Project work feeds the recurring base, keeps engineers sharp, and often carries great margins. The distinction is between projects that orbit managed clients (fine, healthy, expected) and projects that are the business model (a valuation problem at exit).
Moving the mix if you have runway
This is among the highest-return uses of a two-year runway, and the playbook is unglamorous: paper the handshake clients onto real agreements, convert the perpetual T&M regulars to managed contracts at their next renewal or incident, attach contracted security and backup services to the existing base, and let low-value hardware resale shrink as a share of the whole. Owners who spend 18 to 24 months on this routinely add a third or more to their eventual price, which makes it the rare project that out-earns the day job. It’s also the first thing we look at in the pre-sale value work we do with owners who aren’t ready to sell yet.
And if you’re closer than that, know your strict MRR percentage before anyone asks. It’s the second question a buyer will put to you, right after “why now,” and having a precise answer signals more about your business than the number itself does.