The true cost of staying on your old helpdesk

Every team that considers leaving its helpdesk prices the switch: migration hours, retraining, the risk of a rough week. Almost nobody prices the stay. Yet doing nothing has a rate card too — it just arrives twelve line items at a time instead of as one scary invoice.

This piece is the ledger for the stay. No vendor names, no feature debates — only the four mechanisms that make a legacy helpdesk quietly more expensive every year you keep it.

Mechanism 1: renewal creep

The price you signed at is not the price you pay. Mainstream per-seat suites raise list prices at renewal — mid-single digits in a normal year, more when an "AI platform fee" gets folded in — and the generous first-year discount that closed the deal rarely survives into year two.

Run the compound math on a seat that started at $79/month: a 7% uplift at each renewal makes it roughly $90 in year two and $97 in year three. Nothing changed about the product you use; the same inbox now costs 23% more. Multiply by seats and by twelve, and renewal creep alone often exceeds what a full migration weekend would have cost.

Mechanism 2: growth math on per-seat pricing

Per-seat pricing means your tooling bill is indexed to your hiring plan.

A worked example. A team runs 6 agents today at an effective $89/seat — $534 a month. Support scales with the customer base, so two years later it's 11 agents; with renewal creep the seat is now $97 — $1,067 a month. The support tool line doubled, precisely during the growth phase when every dollar of margin was spoken for.

The quieter damage is behavioural: when every hire adds a visible line item, teams delay hiring. Queues stretch, agents burn out, CSAT sags — to optimise a software invoice. Staffing decisions should follow workload, not the vendor's pricing model.

Mechanism 3: add-on stacking

Compare the seat price on the pricing page with your actual invoice. On legacy suites the advertised number is typically 50–60% of reality once the team adds what modern support actually requires: the AI agent SKU, quality assurance, advanced analytics, the second channel pack. Each one is priced per seat, so every add-on multiplies with every hire — the two mechanisms above feed each other.

The pattern to watch at renewal: capabilities that are table stakes elsewhere arriving as paid SKUs on your contract. You're not buying new value; you're re-buying the current decade.

Mechanism 4: the deflection you never turned on

This is the biggest line and the only invisible one — an opportunity cost.

A modern AI agent working from a decent knowledge base resolves 60–70% of routine questions on its own. Staying on a tool where AI is an expensive add-on (or a retrieval bot that answers like a search box) means your humans keep handling that share by hand. The math for a modest team: 2,000 conversations a month, 8 minutes average handle time, 60% deflectable — that's 160 agent-hours every month, or one full-time agent doing work software should absorb. At a loaded cost of $3,500–4,500 a month, the "free" decision to stay costs more than most teams' entire tooling budget.

The switching-cost myth

Against all that stands the one-time cost of leaving, and it's smaller than its reputation. A prepared migration is a weekend of focused work — freeze Friday, import Saturday, test Sunday, cut over Monday — plus a learning dip measured in days. (We published the exact Friday-to-Monday playbook separately.) History, contacts, articles and macros all move; the things that don't move are mostly things that shouldn't.

The honest comparison is a recurring monthly delta against a one-time weekend. Put numbers on a 24-month ledger for the 6→11-agent team above:

• Staying: creeping seats plus one per-seat AI add-on land around $36–40K over two years.
• Switching: the migration weekend plus a flat workspace-priced tool with AI included lands around $8–10K — the switch itself included.

A ~$28K difference is not an optimisation rounding error. It's a hire, a quarter of paid acquisition, or simply runway.

When staying is the right call

The ledger doesn't always point at the door. Staying wins when:

• an active contract carries real early-exit penalties — do the weekend the month before renewal instead;
• a deeply custom workflow genuinely earns money and the new tool can't replicate it yet;
• your prepaid seats run another 18 months — in which case the action item is a calendar reminder, not a migration.

"We're used to it" is not on this list. Familiarity is a training cost, and it's paid once.

Three questions for your renewal call

1. What is our effective per-seat price in year three, all discounts expired, all current add-ons included?
2. Which of the capabilities we use are in the base plan versus separate SKUs, and which SKUs are new since we signed?
3. If AI deflection lets us run with half the seats, what happens to our bill — and to our contract minimums?

If the answers take two weeks and a "let me loop in your account executive", that's an answer too.

Where MoveDesk fits

MoveDesk prices the workspace, not the people: unlimited teammates on every plan, the AI agent included rather than sold back per seat, and migration done for you, free, with both tools running in parallel while you compare. The point of that model is that the ledger above stops compounding — hiring an agent doesn't raise your software bill, and deflection benefits you instead of threatening a seat-count clause.

Run your own numbers in the 14-day trial: import your real tickets, watch what the AI actually resolves, and compare the two columns yourself.

Curious what the ledger says for your team? Start the free trial — full features, no credit card required.