Every support desk is three jobs wearing one name, and a buyer comparing two monthly figures is usually comparing two different answers to the question of which job picks up the phone. Tier 1 restores a known state. Tier 2 diagnoses and configures. Tier 3 engineers and rebuilds. The rate is one blended number covering all three, and of the 29 providers read for this record on 10 September 2026, not one publishes what share of it buys which layer. The tier structure is the thing to ask about, because it decides both the wait and the outcome.
How IT support teams are tiered, and what each tier is allowed to do
Tier 1 works from a runbook: password resets, account lockouts, a stuck print queue, a mailbox that will not open, a laptop taken back to a known good state. It is the layer that restores rather than investigates. Tier 2 is allowed to change things: tenancy settings, group membership, a mail flow rule, a driver that is wrong across a fleet, a firewall rule for a new application. Tier 3 designs and rebuilds: identity, the domain, a failed server, a migration plan, the thing that has never happened before. Ask which tier answers a new call, what happens to a ticket that stalls, and which of the three is on the clock during your working day rather than reachable in principle.
What the types of IT support in a service desk services proposal actually enumerate
Behind the phrase are five separately priced things, and a proposal that bundles them into one line is hiding the joins. The desk itself, quoted per user per month. Remote hands on work, normally inside that fee. Onsite attendance, usually an included count with an hourly rate and travel beyond it. Project work, which is anything the provider classes as change rather than fault, quoted as a fixed fee or a day rate. And the account relationship, the review meeting and the roadmap, which is the line most often promised and least often delivered because nobody bills for it. This record can price only the first of the five: 7 of the 29 providers publish a per-seat figure and none publishes a day rate or a project fee anywhere on its own pages.
Tier 2 is where the queue stalls and the money goes
Tier 1 is trainable and cheap to staff, and tier 3 is rare and usually billed as a project, so the layer that determines how a contract feels in practice is the middle one. It is also where a thin provider is thin. Useful questions have numbers in them: how many tier 2 engineers cover your time zone across the whole client base, whether tier 1 and tier 2 sit in the same business or the desk is white labelled from a third party, whether escalation is a warm handover with the engineer still on the call or a queue transfer that starts again, and whether escalating a ticket resets the response commitment. A firm that cannot answer the headcount question is usually answering it by not having one.
Why the first line decides how IT service support is judged
The cheapest layer to deliver is the only one most of the business ever meets, which is why contracts that are technically sound get cancelled on the experience of the first call. The details that decide it are unglamorous: whether the phone, the portal and the mailbox reach the same people, whether a caller has to restate the estate every time, whether the desk knows which machine in accounts must not be touched during month end. Ask for the name of the person who runs the desk, ask whether first line is employed or subcontracted, and ask to call the published number once before signing anything. A minute on hold as a stranger tells you more than a reference call the provider arranged.
The one measurement to ask for, and what the published rates hide
First response time is the metric the trade volunteers and it is the easiest to satisfy without helping anybody: an automatic acknowledgement and a triage touch both stop that clock. The number that describes real skill on the front line is the share of tickets closed by the person who first picked them up, asked for over the last quarter and for businesses of your size, together with how a reopened ticket is counted. Then look at what is published. In this record the lowest figure is Electric's $10/user/mo and Ntiva states Starting at $99 per user, with a per-seat median across the seven priced providers of $99.50 a user a month and Be Structured Technology Group's schema markup carrying $125-$300 per user, per month. A rate near the bottom of that spread is not carrying a resident tier 3, and a rate near the top should be asked to name the engineers it does carry.
Questions people ask about IT support desk
What is tier 2 IT support?
The layer allowed to change configuration rather than restore a known state: tenancy and mailbox settings, group membership, mail flow rules, fleet-wide driver and application faults. Tier 1 restores, tier 2 diagnoses and configures, tier 3 rebuilds and designs.
Which tier answers when I call?
Almost always tier 1, and that is normal. What matters is what fraction of calls never needs to leave it, and how a ticket reaches tier 2 when it does: a warm handover with the caller still on the line, or a transfer to a queue that starts from nothing.
Do providers price managed support services by tier?
Not in this record. All seven providers that publish anything publish a single blended per-seat rate, six per user and one per workstation, with no breakdown by tier. The split has to be asked for, and it is a staffing question rather than a pricing one.
Is an IT support MSP the same as buying managed support?
In pricing terms usually yes: one monthly per-seat fee wrapping all three tiers. The difference worth checking is who owns escalation past tier 3, meaning who holds the vendor relationship for your line of business application and the global administrative credential for your Microsoft 365 tenancy.