A monthly IT number is only useful if you can point to the catalog behind it. Mid-market teams pay for coverage, hygiene, and a named escalation path — not a slogan.
Quotes that say “per user” and nothing else are how mid-market shops pay twice: once in the retainer and again as project work. Pricing is a scope document with a unit attached. If you cannot explain the unit to finance on one page, you do not have a price — you have a conversation that will reopen at renewal.
Service design lives on managed IT services. Published entry points are on pricing and plans. Cloud waste that sits next to the ops stack is a FinOps consulting conversation, not a hidden line inside the MSP invoice.
The model is how the unit scales. It is not a quality claim.
Finance likes this when headcount and endpoints move together. It fits poorly when you have shared workstations, kiosks, lab devices, or a large server and network footprint that has nothing to do with how many people log in.
Clearer when the work is patch, backup, and monitoring on machines. You still need a rule for servers, network gear, and cloud accounts — those are not “laptops with a different SKU.”
A named catalog (monitor + patch + backup + after-hours, then add-ons). Useful when the tiers are written. Useless when “Gold” is just a higher number and the same queue.
A retainer for the operating layer, plus a rate or a block for work outside the catalog. Honest when the catalog exists. A trap when everything interesting is “out of scope.”
Pick the model by what you can count and what you can exclude. Then write both lists. That is the price.
Mid-market retainers that survive a finance review usually name these as in-scope:
Usually excluded, and should stay excluded unless you buy them as projects: net-new sites, major version upgrades, application development, and anything that needs a design workshop. If those sit inside the monthly number with no unit, you will argue about them later.
Cost is not “MSP is cheaper” or “co-managed is cheaper.” The models buy different ownership.
You are paying for the provider to own the queue, the tooling, and most of the environment. Your staff become requestors. The monthly figure should cover that full operating layer. It looks simple. It gets expensive when your applications team still needs admin rights the MSP was not scoped to share.
You keep identity, application owners, and business-facing IT. The partner takes monitoring, patch, backup, after-hours, and overflow. You are not buying “fewer people.” You are buying coverage on the layer you cannot staff 24/7. The monthly number should be smaller than a full MSP only if your team is really keeping the rest. If they are not, co-managed is just an under-scoped MSP.
Staff augmentation without an SLA is not a third pricing model. It is contracting. When the contractor is out, coverage is out. Price it as a project or a named backfill — not as managed IT.
These are the lines that do not appear on the first slide and show up in month three:
We will not invent a typical monthly range or a savings percentage. Your number is the catalog times your inventory. If a vendor leads with a range and cannot show the catalog, treat the range as decoration.
Discovery is the paid scoping step. It is not a monthly quote and it is not an implementation. For $99 you get an IT ops map: systems, owners, coverage gaps, and what belongs in a retainer versus a project. That map is what a later SOW should quote against.
Book it on Discovery. Published commercial paths sit on pricing and plans. The managed IT landing is the service page that quote should match.
Discovery maps what your team should keep and what belongs under a written catalog — before anyone invents a monthly number.
Also see pricing · managed IT · FinOps