IT budgets go wrong in predictable ways. A server dies and becomes an unplanned purchase. Software subscriptions creep up seat by seat. The managed service fee is compared against last year's break-fix invoices without counting the downtime those invoices came with.
This post explains the two kinds of IT spend, how providers price managed services and why, and a simple structure for a yearly IT budget that a business owner can read in ten minutes. Check the tax treatment with your accountant.
Capex and opex in IT terms
Capital expenditure is money spent on something the business owns and uses for years: servers, network equipment, laptops bought outright, a cabling install. It is typically depreciated over its useful life on the books. Operating expenditure is money spent to run the business month to month: subscriptions, managed service fees, internet circuits, leased equipment, support contracts.
Over the past decade most IT spend has shifted from capex to opex. Email became a subscription, servers became cloud instances or leases, and support became a monthly fee. The shift is not automatically good or bad. Opex is predictable and scales with headcount. Capex can be cheaper over a long life for stable workloads. A good budget uses both on purpose.
- Capex: servers, switches, firewalls, laptops purchased, cabling, racks and UPS
- Opex: Microsoft 365 or Google Workspace, managed IT fees, cloud hosting, circuits, licences, warranties
- Either: hardware can be bought (capex) or leased (opex); the choice is about cash flow and refresh cadence
- Hidden opex: the internal time spent on IT by people whose job is not IT
How managed IT is priced
Per-seat pricing charges a monthly fee per user and covers their devices, their accounts and their support. Per-device pricing charges per managed endpoint and server, which suits companies with shared workstations or many devices per person. Flat-fee pricing sets one monthly number for the whole environment based on a documented scope, and is reconciled when the scope changes.
Providers build these from the same inputs: the cost of the tool stack per device (RMM, security, backup, documentation), the expected support hours per user per month, the after-hours coverage, and a margin. The tier names vary but the difference between tiers is usually what is bundled: whether security tooling, backups, projects or after-hours coverage are inside the fee or billed separately. Ask what is inside and what is not, then compare the total, not the headline number.
- Per seat: simple, scales with headcount, best for one-device-per-person offices
- Per device: suits shared workstations, kiosks, and server-heavy environments
- Flat fee: predictable, needs a clear scope and a reconciliation rule
- Look inside the tier: security stack, backup, projects, after-hours, hardware procurement
- Ask how mid-month adds and removals are billed, and when a count is reconciled
Building a yearly IT budget
A usable budget has five lines and a refresh schedule. The lines are: recurring services (managed IT, subscriptions, circuits), planned hardware refresh (from the asset inventory), planned projects (from the roadmap agreed at the quarterly review), a contingency for the unplanned, and security and compliance spend if it is separate. Each line has a monthly and an annual figure and a note on what drives it.
The refresh schedule is what makes the hardware line honest. Take the inventory, sort by policy age and warranty end, and spread the replacements across the year so the spend is even and the work is manageable. The contingency line covers what the schedule cannot predict; size it from the last two years of unplanned spend, and if that number is large, it is a sign that the refresh schedule needs to be more aggressive.
- List recurring services with the renewal date and the expected change in seats
- Pull the refresh list from the inventory and spread it across quarters
- Add projects from the roadmap with a cost band and a quarter
- Set contingency from recent unplanned spend
- Review the budget at each quarterly review and adjust the remaining quarters
Comparing managed IT against the alternative
The honest comparison is not the managed fee against the break-fix invoices. It is the managed fee against break-fix invoices plus the cost of downtime, plus the cost of the internal time spent on IT, plus the security incidents that a maintained environment would have prevented. Nobody can price the last two exactly, but everyone can list last year's outages and the hours they cost.
Put that list beside the two budgets and the decision is usually clear one way or the other. Where the environment is small and stable, break-fix with a good documentation baseline can be fine. Where the business depends on IT working every day, the managed fee is buying the absence of a list. If you would like a budget template built around the five lines above, RackLedge can share one.
Frequently asked questions
Should we lease or buy laptops?
Leasing spreads the cost and forces a refresh cadence, which keeps the fleet healthy. Buying is cheaper over a long life if the business keeps devices well past the lease term. Decide based on cash flow and whether a forced refresh is a feature for you.
How much of the IT budget should be contingency?
Enough to cover a typical unplanned failure at your scale, based on the last couple of years. If contingency is being spent every year, move that spend into the planned refresh line where it belongs.
Is per-seat pricing fair when some users barely use IT?
Most providers handle this with a reduced rate for light users or by counting devices instead. Ask. The goal is a unit that both sides can count without arguing.
Takeaway
Know which spend is capex and which is opex, and choose each on purpose. Understand what is inside a managed IT tier before comparing prices. Build the budget from five lines and a refresh schedule pulled from the inventory, and review it every quarter. The budget that survives is the one that predicted the server refresh a year before the server failed.