Switching payroll mid-year: the cost that is in no comparison table

This page exists because a comparison table that never mentions switching is an argument for switching, and we would rather not make that argument by omission.

What does switching actually involve?

Because payroll carries history, four things have to move with you. Not one of them appears on a pricing page.

Year-to-date figures have to move. Every employee's earnings, tax withheld, and deductions so far this year have to be entered into the new system accurately, because the new system will calculate the rest of the year from them. Get one wrong and the error compounds through every subsequent pay run and shows up at year end.

Filings have to be reconciled. Both providers will have filed part of the year. Which one files the quarterly return covering the transition, and on what basis, is a question with a right answer that neither vendor will necessarily volunteer.

Every downstream connection breaks too — accounting software, time tracking, benefits providers, pension schemes, whatever pulls payroll data. Each needs reconnecting and each needs checking once.

Someone has to learn the new system before a real pay run. Payroll is unforgiving in a way most software is not — the deadline is fixed, the recipients notice immediately, and the regulator notices later.

Why is mid-year worse than year-end?

Because at year-end there are no year-to-date figures to carry across. Starting a new provider on the first pay run of a tax year removes the largest single source of error and the reconciliation question along with it.

In the United States that means January. In the United Kingdom it means the sixth of April. Any vendor's sales team knows this, which is why introductory discounts cluster around those dates — and why a discount offered in the middle of a tax year is being offered against a genuinely worse time to move.

If you are reading a comparison in August, the honest answer is often "this is the right choice, and the right time to make it is five months from now."

How large does the saving have to be?

It depends on a cost you have to estimate yourself, because nobody else can see it. The saving has to be large enough to clear that cost. What we can do is set out the shape of it.

Suppose moving costs two days of someone's time. At a fully loaded cost of $400 a day, that is $800. A saving of $40 a month clears it in twenty months. A saving of $101 a month — the gap between the cheapest and most expensive mainstream US options at ten employees is around that, at $57 against $158 — clears it in eight.

That arithmetic is not a recommendation, and the $400 figure is an illustration rather than a finding; substitute your own. What it shows is the shape of the answer: small savings do not justify mid-year moves; large ones do, and the boundary sits somewhere most buyers do not check before starting.

When is switching clearly worth it?

Because the strongest reasons to switch are not about price at all. Three cases come up repeatedly and none of them are on a pricing page.

Your current provider cannot do something you now need. Multi-state filing after a remote hire is the classic example. If the alternative is doing it manually every month, the comparison is not between two monthly fees.

You are on a quote-only contract and have never tested it. Six of the largest US providers and five UK ones publish no prices at all, and a contract signed three years ago without a benchmark is very likely above market. Getting published figures for your headcount — which is what the calculator is for — costs nothing and gives you a number to take back to your account manager. Renegotiating is cheaper than moving.

Your headcount has crossed a band. Vendors that price in bands step sharply. Crossing from Breathe's one-to-ten band into eleven-to-twenty raises the bill by 83 percent in a single hire, and a band boundary is a natural moment to re-run the comparison rather than absorb the step.

What we do not know

Whether switching costs two days or ten. Whether your provider will make the year-to-date export easy or awkward. Whether the new one's onboarding is good. We have not run payroll on any of these products and we have not surveyed anyone who has, so anything we said about that would be invention rather than information.

What this site does hold is the published price side of the equation, checked and dated, with the source for every figure. The other side you have to supply.

Who this is not for

This is not for a business setting up payroll for the first time. None of it applies — there is nothing to migrate, no reconciliation, no band to cross. Go straight to the vendor list and compare on published cost, because for you the published cost is close to the whole cost.

It is also not for anyone whose payroll is currently going wrong. If filings are late or wrong, the cost of staying is not a monthly fee, and a comparison table is the wrong tool entirely.

Sources

The price spreads quoted (US published plans at ten employees ranging from $57 to $158; Breathe's £24 and £44 bands) come from the named vendors' own pricing pages, read between 2026-08-20 and 2026-09-06, and are recorded with source URLs in /data.json. The daily-cost figure used in the break-even illustration is an assumption stated as such, not a measurement — it is not drawn from any source and should be replaced with your own. How we handle figures we cannot source is set out on the method page.