What Makes a Great MSP?

Ask any MSP what makes them different and you will hear the same six things, usually in the same order. SOC 2. Twenty-four-seven monitoring. Tested backups. A named engineer on your account. Clear SLAs. An offboarding clause that doesn't quietly hold your data hostage. I built a checklist along exactly those lines a while back, because those questions are still the fastest way to separate a provider worth talking to from one that isn't. They work. Every business evaluating an MSP should ask them, in writing, before signing anything.

Here is the part that took me longer to see. Two providers can answer every one of those questions correctly, hand you the same attestation letter, quote you almost the same price, and still turn out to be completely different companies three years into the relationship. One of them is doing in year three exactly what they said in year one. The other has quietly become someone else — the technicians rotate, the reporting thins out, the SLA turns into a number in a slide nobody checks against reality — and you don't find out until the week something breaks. The checklist got both of them in the door. It never told you which one you were actually hiring.

I've sat on enough of these evaluations to stop being surprised by it. The scorecard and the year-three reality belong to two different documents, and which one you actually signed up for is usually something you only find out after the fact.

The checklist gets you to average

The reason the checklist stopped being a real differentiator is that the whole market learned it at the same time. Kaseya's 2026 State of the MSP report has 71% of providers naming customer acquisition as their single biggest challenge this year, ahead of talent, ahead of margin. Growth is harder to find, the buyers on the other end of every call have read the same three vetting guides, and RFPs across the industry now ask for the same six proof points as a matter of course.

That is a good thing for the buyer in one sense. It also means the checklist has become table stakes rather than a filter. Passing it tells you a provider is not obviously reckless. It does not tell you they are good.

The consolidation happening underneath the industry pushes the same way. A 2026 tally of MSP transactions counted 169 deals in 2025, with private equity involved in 69% of the ones where the buyer was disclosed. One acquirer, Evergreen Services Group, closed 47 MSP acquisitions in a single year; its platform Lyra Technology Group crossed $1 billion in annual recurring revenue and its hundredth cumulative acquisition within the same stretch. A meaningful share of the shops you call for a quote now trace back to a small number of PE-backed platforms, running the same tooling, hitting the same compliance bar, because a sponsor several ownership layers up requires it before the next sale.

The checklist is converging right along with the market that produces it.

None of this makes private equity the villain of the piece, and I want to argue against my own drift here for a paragraph, because the honest version is more useful than the tidy one. Some of the most disciplined MSPs I have reviewed are PE-backed platforms, and for a mundane reason: a sponsor preparing a company for its next sale wants uniform tooling, documented processes, and a clean security story across every acquired branch, because that is what the next buyer will diligence. A ten-person shop bootstrapped on a founder's savings often cannot afford the SIEM, the 24/7 desk, or the compliance program that a platform can deploy across forty locations at once. Scale is not automatically the enemy of quality. The failure mode isn't the ownership structure. It's what happens to the specific relationship you signed up for in the handoff — whether the engineer who knew your environment is still the one answering the phone, or whether you've been quietly reassigned to a pool.

Where the difference actually lives

So if the checklist is necessary but no longer sufficient, what is the rest of it? In every relationship I've watched hold up for years, the answer comes down to something less quotable than a certification: the story the contract tells, the story the delivery tells, and the story the client believes, all stay the same story, and they stay that way without the client having to go check.

That shows up in small, specific behaviors more than in any single capability. A good provider tells you about the bad week before you find it yourself. When a backup job has been failing quietly or a patch window got missed, the report that lands on the first of the month says so, with a plan attached, rather than staying silent until a restore fails and forces the conversation. A weaker provider waits to be asked, and answers only the question you actually put to them, which is a very different thing from disclosure.

A good provider treats "delivered" as something you can point to, not something you can say. Patching is a dated, logged event you can pull up on a screen, not a status on a slide. A backup is not proven by a green dashboard tile; it is proven by a restore that was actually run, timed, and documented, on a cadence the client can see. I have sat through enough vendor calls to know the gap between "we handle that" and someone actually screen-sharing the evidence, unprompted, before you thought to ask for it. That gap is most of what you are paying for.

And a good provider keeps the specific person constant, or tells you the moment they can't. The named engineer on the contract is only worth something if you're told when that person leaves the account, gets promoted, or gets folded into a shared queue after an acquisition. The silent version of that same event — where the name on the org chart quietly stops meaning anything — is one of the most common ways a fine relationship turns into an average one, and it rarely shows up as a single bad day. It shows up as a slow decline nobody flags until the renewal conversation gets tense.

What I would do about it

Ask the checklist questions, but treat a clean scorecard as a floor, not a verdict. Passing SOC 2, MFA, and a documented restore test gets a provider onto your shortlist. It does not tell you which one on that shortlist is going to still be the same company in year three.

Ask who owns the account by name, and ask what happens to your service the day that person leaves or the company gets acquired. A provider with a real answer has thought about this before you asked. A provider that looks surprised by the question has not.

Ask to see a real report, not a sample one, from a real client relationship that's run for at least two years. What a provider chooses to disclose without being asked tells you more than anything on their proof-of-concept deck.

If your current provider has been through a change of ownership, treat the next ninety days as a checkpoint, not an inconvenience. Ask directly whether your named contacts, your SLAs, and your escalation path changed, and get the answer in writing rather than assuming continuity because nobody told you otherwise.

And revisit the relationship on a schedule, not just at renewal. The gap between what you're paying for and what you're actually getting rarely opens all at once. It drifts, a little every quarter, until the year you finally look closely and find out how far apart the two pictures had gotten.

A great MSP is not the one with the best answers on the call before you sign. Plenty of average ones have learned to give those same answers well. A great MSP is the one still giving you the true version of those answers, unprompted, three renewals from now — and the only way to know which kind you hired is to keep checking after the ink is dry, not just before.

Sources

*Kaseya, 2026 State of the MSP Report (2026), on customer acquisition as MSPs' top-cited challenge; CT Acquisitions, Private Equity MSP (2026) market guide, on 2025 MSP deal volume, private-equity deal share, and Evergreen Services Group / Lyra Technology Group acquisition activity. Figures reflect public reporting available as of mid-2026.*

A great MSP is not the one with the best answers before you sign. It's the one still giving you the true version of those answers, unprompted, three renewals from now.