Three managed service provider proposals land on your desk. All three promise 24/7 support, strong security and a smooth transition. One costs $4,000 a month less than the others. Guess which one your CFO asks about first.
MSP evaluation criteria are the specific, weighted factors a buying committee uses to compare managed IT and security providers on the same terms: security depth, compliance expertise, service levels, strategic guidance, breadth of services, scalability, onboarding and pricing transparency. Each criterion gets a weight before anyone reads a proposal, each provider gets a 1–5 score backed by evidence, and the weighted total decides the shortlist.
This guide walks through how to compare MSPs with a weighted scoring model: the eight criteria, what weak and strong answers look like, how to set the weights, how to line up pricing, and how to run the scoring session so the result holds up when leadership asks how you got there.
Why choosing an MSP on gut feel or lowest price backfires
Most MSP selections go wrong the same way. Each evaluator reads the proposals in their own order, weighs what they personally care about, and walks into the decision meeting with a favorite. The loudest opinion wins, or the cheapest bid does, because price is the one number everyone can compare.
The problem is that proposals hide their differences in the details:
- A response SLA is not a resolution SLA. “We respond within 15 minutes” says nothing about when your outage ends.
- Low monthly fees move cost elsewhere. Onboarding fees, tooling licenses and hourly rates for after-hours work can flip the three-year math.
- “24/7 security” can mean an alert inbox. Unless the proposal names who staffs the SOC and who has authority to contain a threat, you’re buying a promise.
The stakes are real. According to IBM’s Cost of a Data Breach Report 2026, the global average cost of a breach reached $4.99 million, a 12% increase over the prior year and a record high. Saving $48,000 a year on a provider that can’t contain an incident is not a saving.
A scorecard fixes the process problem: one set of criteria, one set of weights and one place to record why every score was given.
Many organizations combine them: an RFI to narrow a long list to four or five qualified providers, then an RFP to that short list. That saves bidders from writing full proposals they have little chance of winning, and saves you from reading them.
Not ready to score yet? Start with the right questions.
How does a weighted scoring model work for vendor selection?
A weighted scoring model is a decision method that assigns each evaluation criterion a percentage weight based on its importance, scores every vendor on each criterion, and multiplies score by weight to produce one comparable total. The weights add up to 100%, so a provider can’t win on volume of features alone; it has to win on the things you said matter most.
The math fits on one line:
Weighted total = sum of (criterion weight × average evaluator score) across all eight criteria
If security depth carries 20% and a provider averages 4.0 across three evaluators, that criterion adds 0.80 points to its total. Repeat for all eight criteria and every provider ends up with a score out of 5.00.
The model does two things a gut-feel review can’t. It forces the committee to agree on priorities in advance, and it turns a disagreement about a vendor into a narrower, answerable question: “Why did you score their onboarding a 2 when I scored it a 4?”
The 8 MSP evaluation criteria: what weak and strong look like
These eight criteria cover what actually determines whether an MSP relationship works three years in. Score each from 1 to 5, and use the descriptions below as your anchors: a weak answer is a 1, a strong one is a 5, and “meets the requirement as stated” is a 3.
| Criterion | What weak looks like (1) | What strong looks like (5) | Evidence to request |
|---|---|---|---|
| 1. Security depth | No true 24/7 monitoring, or a SOC quietly subcontracted with no clarity on who acts. Antivirus and backups pitched as a security program. | In-house 24/7 SOC with MDR/XDR and authority to contain. Reports detection and response times. Shares a tested incident response plan. | SOC 2 Type II report, redacted incident report, who staffs the SOC and where |
| 2. Industry and compliance expertise | No clients in your vertical. Generic compliance claims. Won't sign a BAA. | Has supported clients through your specific audits (HIPAA, PCI DSS, CMMC, SOC 2). Maps controls to your framework and helps with cyber-insurance questionnaires. | Industry client count, audit examples, sample BAA |
| 3. Service levels and accountability | Best effort" or response-only SLAs. No reporting. No remedies when targets are missed. | Publishes 12 months of actual SLA attainment. Resolution targets backed by service credits. Can walk through a 2 a.m. P1 outage step by step. | 12-month SLA attainment report, escalation matrix, remedies clause |
| 4. Strategic alignment | No named strategic resource. Business reviews are ticket counts. | Named vCIO and/or vCISO. Shares a sample roadmap, budget and risk register. Cadence written into the contract. | Sample QBR deck, 12–24 month roadmap, advisor bios |
| 5. Breadth of services | Covers part of your scope. Key services subcontracted and undisclosed. | Help desk, network, cloud and security delivered by one team that owns the outcome. Flags gaps you didn't list. | Scope matrix against your requirements, subcontractor list |
| 6. Scalability and footprint | Remote-only, or no on-site coverage at some of your sites. Pricing doesn't scale. | Named field coverage for every location, a playbook for new sites and growth pricing in the contract. | Coverage map vs. your locations, multi-site references |
| 7. Onboarding approach | Vague timeline, no named team. The provider keeps your documentation and admin credentials. Steep exit fees. | Detailed 30/60/90-day plan with risks and rollback. You own all documentation and credentials. Exit assistance is contractual. | Onboarding project plan, documentation sample, exit clause |
| 8. Pricing transparency | Pricing doesn't map to line items. Tooling hidden in bundles. Uncapped escalators. | Every line priced or marked included. Capped escalator. Lowest or near-lowest 36-month cost for equivalent scope. | Completed pricing sheet, rate card, sample invoice |
One rule makes the rubric work: if no one can point to evidence for a 4 or a 5, the score is a 3. Write down what you’d need to see to raise it, and ask the provider for it.
Some requirements shouldn’t be scored at all. A 24/7 SOC, a signed BAA or a current SOC 2 Type II report may be pass/fail for your organization. Treat those as gates before scoring, so a provider can’t make up for a missing must-have with points elsewhere. For a deeper look at what to measure once a provider is in place, see our guide to metrics to evaluate your MSP.
How to set the weights (and why you set them before reading proposals)
Set your weights before anyone opens a proposal, and lock them. Once evaluators have seen the bids, every weight becomes a vote for a favorite: the person who liked the cheapest bid suddenly thinks pricing deserves 30%. Weights agreed in advance keep the committee honest with itself.
Start from your situation, not a blank page. These four presets each total 100% and cover most mid-market evaluations:
| Criterion | Balanced | Compliance-heavy | Multi-site growth | Security-first |
|---|---|---|---|---|
| Security depth | 20% | 20% | 15% | 35% |
| Industry and compliance expertise | 15% | 25% | 10% | 15% |
| Service levels and accountability | 15% | 15% | 15% | 15% |
| Strategic alignment | 10% | 15% | 10% | 10% |
| Breadth of services | 10% | 5% | 15% | 5% |
| Scalability and footprint | 10% | 5% | 20% | 5% |
| Onboarding approach | 10% | 5% | 5% | 5% |
| Pricing transparency | 10% | 10% | 10% | 10% |
- Balanced fits when no single risk dominates the decision.
- Compliance-heavy fits healthcare, financial services, legal and anyone facing audits or cyber-insurance questionnaires.
- Multi-site growth fits organizations adding locations or acquiring practices.
- Security-first fits after an incident, a failed audit or an insurance non-renewal.
Notice that pricing transparency holds at 10% in every preset. That’s deliberate. Price still matters, but it enters the decision as one criterion scored on the full contract cost, not as a tiebreaker that overrides everything else.
Free Template: Get the MSP Proposal Scorecard
How to compare MSP pricing line by line
Compare MSP pricing on 36-month total cost, not the monthly fee. A proposal with the lowest monthly number can easily cost more over the contract once onboarding, tooling, hourly labor and annual escalators are added in.
To get a fair comparison, put every bid on the same lines:
- One-time costs: onboarding and transition, discovery and documentation, security baseline remediation, tool deployment, migration.
- Recurring services: service desk, after-hours support, endpoint and network management, backup and DR, 24/7 SOC and MDR, vCIO and vCISO time.
- Tooling and licenses: EDR, RMM, backup storage, email security, MFA. These are often bundled into one line or left out entirely.
- Hourly rates: after-hours support, out-of-scope work, emergency on-site dispatch and project engineering. Multiply each by your realistic monthly estimate.
- Annual items and terms: pen tests, risk assessments, true-ups, the escalator percentage, contract term and early-termination terms.
Treat every blank line as a question, not a zero. “Not quoted” usually turns into a change order in month four. Ask each provider to confirm in writing whether a blank line is included or excluded, then score pricing transparency on the answers. If you haven’t sent your RFP yet, asking every provider to price on these same lines will save you a week of spreadsheet archaeology.
Worked example: when the cheaper MSP loses the weighted score
Here’s how the numbers play out for a 150-user organization comparing two finalists with the Balanced weights. The figures are illustrative, but the pattern shows up in evaluation after evaluation.
Provider A quotes $18,500 a month. Provider B quotes $22,500. On the sticker, A saves $4,000 a month. But A charges $22,000 for onboarding, bills after-hours and out-of-scope work hourly (about $2,800 a month at realistic volumes) and has an uncapped 5% annual escalator. B includes most of that work, charges $8,000 for onboarding and caps its escalator at 3%.
Over 36 months, A costs about $828,000 and B about $865,000. The $144,000 headline gap shrinks to roughly $37,000, or about $7 per user per month.
Then three evaluators score both providers independently:
| Criterion | Weight | Provider A avg. | A weighted | Provider B avg. | B weighted |
|---|---|---|---|---|---|
| Security depth | 20% | 2.67 | 0.53 | 4.33 | 0.87 |
| Industry and compliance expertise | 15% | 3.00 | 0.45 | 4.00 | 0.60 |
| Service levels and accountability | 15% | 2.33 | 0.35 | 4.33 | 0.65 |
| Strategic alignment | 10% | 2.00 | 0.20 | 4.00 | 0.40 |
| Breadth of services | 10% | 3.33 | 0.33 | 4.00 | 0.40 |
| Scalability and footprint | 10% | 3.00 | 0.30 | 3.67 | 0.37 |
| Onboarding approach | 10% | 2.67 | 0.27 | 4.33 | 0.43 |
| Pricing transparency | 10% | 4.00 | 0.40 | 3.33 | 0.33 |
| Weighted total (out of 5.00) | 100% | 2.83 | 4.05 |
Provider A wins on price and loses almost everywhere else. Its SOC is subcontracted with no named escalation owner, its SLA covers response time only, and its onboarding plan is a one-page timeline. Provider B finishes 1.22 points ahead.
The point isn’t that the expensive option always wins. It’s that the committee can now tell the CFO exactly what the extra $37,000 buys: a resolution SLA with service credits, an in-house SOC with authority to contain threats, and a named vCISO. That’s a decision leadership can approve, or push back on, with the facts in front of them.
How to run the MSP scoring session
A good scoring session takes 60 to 90 minutes per round and ends with a ranked shortlist everyone can defend. Run it in this order:
- Confirm the gates. Remove any provider that fails a pass/fail requirement before discussing scores.
- Score independently first. Each evaluator scores all eight criteria on their own, using the rubric, before the meeting. No comparing notes. Independent scores surface real differences in how people read a proposal.
- Log the evidence. Every score cites its source: a proposal page, a reference call, a SOC 2 report, a live demo.
- Find the spreads. Flag any criterion where evaluators differ by 2 points or more. That’s where the conversation belongs, not on criteria everyone already agrees on.
- Discuss, then let evaluators adjust their own scores. No one overrides anyone else’s number. If the discussion surfaces new evidence, people will move on their own.
- Send follow-up questions. Any criterion stuck at 3 for lack of evidence becomes a specific request to the provider.
- Summarize for approvers. Give the CFO and COO one page: weighted totals, the top three differences between finalists, 36-month cost, key risks and the committee’s recommendation.
Include at least one person from finance, one from IT and one from operations. Each will catch different gaps, and the approvers will trust a result their own people helped build.
MSP evaluation FAQ
What are the most important MSP evaluation criteria?
For most mid-market organizations, security depth, industry and compliance expertise, and service levels carry the most weight, together about half of the decision. The right mix depends on your risks: a healthcare group facing audits weights compliance higher, while a company opening new sites weights scalability and on-site footprint higher.
How many MSPs should we compare?
Three to five is the practical range. Fewer than three gives you little leverage or comparison; more than five overloads evaluators and stretches the timeline. Use pass/fail gates to trim a long list to a shortlist before full scoring.
How many MSPs should we compare?
Three to five is the practical range. Fewer than three gives you little leverage or comparison; more than five overloads evaluators and stretches the timeline. Use pass/fail gates to trim a long list to a shortlist before full scoring.
What is an MSP scorecard?
An MSP scorecard is a structured template that lists weighted evaluation criteria, a scoring rubric, a column for each evaluator and a pricing comparison, so a committee can score every provider the same way and produce one defensible total.
Should the lowest-priced MSP proposal win?
Not by default. Price is one criterion, best judged on 36-month total cost for equivalent scope. The lowest monthly fee often hides onboarding, tooling or hourly charges, and a weak security or SLA score can cost far more than the difference.
How long does an MSP evaluation take?
A focused evaluation typically runs four to eight weeks from RFP to signed contract: about two to three weeks for providers to respond, one to two weeks for scoring and reference calls, and the rest for final presentations and contract review.
Choose the MSP you can defend, not just the one you can afford
The right MSP evaluation criteria won’t make the decision for you, but they make it visible. When every provider is scored on the same eight criteria, with weights set in advance and evidence behind every number, the committee can show exactly why the winner won. That’s the version of the decision your CFO, your board and your future self will thank you for.