Xero has announced a price rise for UK subscribers that takes effect on 1 September 2026. If you run a practice, you have probably had the email, and possibly a few more from clients asking what it means for them. The temptation is to do one of two things: shrug and absorb it, or start pulling client books off the platform in a hurry. Both are usually mistakes.
This is a guide to the third option: treat the rise as a prompt to do a costing exercise most practices should do once a year anyway, and then make a deliberate decision. Most of what follows is useful whether you stay on Xero, move, or do a bit of both.
What Xero has announced
According to Xero's pricing update, the following UK business plan prices apply from 1 September 2026, all excluding VAT:
- Ignite moves from £16 to £18 a month, a rise of 12.5 per cent.
- Grow moves from £37 to £39 a month, about 5.4 per cent.
- Comprehensive moves from £50 to £55 a month, 10 per cent.
- Ultimate moves from £65 to £70 a month, about 7.7 per cent.
- Simple stays at £7 a month.
The changes apply to new and existing UK subscribers. Xero has also announced that the multi-organisation discount will no longer be applied from the same date. For a practice holding many organisations under one subscription, that removal can matter more than the headline increases, so check your subscription page and work out exactly what discount you receive today before you estimate the new bill. Xero has said that other existing discounts and promo codes continue until they expire. And as with any published figure, confirm the current numbers on Xero's own pricing pages before you build a budget on them.
Two things are worth saying in fairness. First, this follows a rise in September 2025, so some plans have moved twice in twelve months, and it is reasonable for practices to feel the compounding. Second, Xero remains a mature, well-supported ledger with a deep app ecosystem and a large pool of bookkeepers who already know it well, and the company attributes the increase to continued product investment. None of that is spin. It is exactly the value you should weigh, deliberately, in the exercise below.
The number that matters is whole-stack cost per client
A ledger subscription is rarely the whole software bill for a client. In most practices the real stack is the ledger, plus a document capture tool, plus AI add-ons, plus usage-metered extras such as payroll per employee or expenses per user, plus anything billed per document or per submission. A £2 rise on one line of that stack tells you very little on its own.
The exercise takes an hour. Pull one month of software invoices that relate to client work: ledger subscriptions, capture tools, AI features, per-user add-ons, metered charges. Add them up and divide by the number of active clients. Then do it again by segment, because a payroll client with expense claimants costs several times what a dormant company costs, and an average across the two hides both.
Imagine a hypothetical 40-client firm with 25 clients on Grow, 10 on Comprehensive and 5 on Ignite. Its ledger cost today is £1,505 a month. From 1 September that becomes £1,615, a rise of £110 a month, or £2.75 per client. Now add what the firm pays per client for capture and AI features, plus the metered lines. In some stacks those lines together rival or exceed the ledger line, and the metered ones drift upwards quietly as document volumes grow. The September rise is the visible part of the bill. The exercise often finds that the invisible part is bigger.
While the numbers are out, check plan fit. Practices accumulate plan levels the way they accumulate old engagement letters: a client put on Comprehensive two years ago for a feature nobody now uses may be fine on Grow or Ignite. Rightsizing plans and cancelling unused add-ons is often worth more than the rise costs, and it requires no migration at all.
If AI coding or capture tools are part of the stack you are reviewing, price is the second question, not the first. The first is how the tool behaves when it is not confident about a figure, which is the subject of review-first versus silent automation.
Questions to answer before you move anything
If the exercise makes you want to look at alternatives, good. Do it with a checklist rather than a grievance. These questions decide whether a move is safe, whatever you might move to.
Where does the data go if you leave?
Ask for the full general ledger with journals and dates, not just a closing trial balance. Ask what happens to source documents and attachments, and to the history of who changed what. Ask what format the export takes and whether the receiving system can import history or only opening balances. A bookkeeping file's value is largely its evidence trail, each entry linked to the document behind it, and a move that strands five years of that behind a cancelled login has a real cost even if no one invoices you for it. What a well-evidenced file looks like is a subject of its own: see the audit quality of a bookkeeping file.
Will your digital links survive the move?
For VAT-registered clients, Making Tax Digital requires digital links through to the return, and a mid-quarter cutover is where those links most often break. Plan any cutover at a VAT quarter boundary, keep the old system's records exportable until the enquiry window has passed, and document how the balances transferred. The calendar is also more crowded than it used to be: since 6 April 2026, sole traders and landlords with qualifying income over £50,000 (measured on gross income, not profit) are in MTD for Income Tax with quarterly updates, and the threshold is due to fall in the coming years. Quiet windows in which to move a client are scarcer than they were. There is a fuller treatment in the MTD VAT 2026 practice checklist.
What does retraining actually cost?
Count the hours honestly: chart of accounts mapping, bank rules rebuilt, invoice templates, report layouts, the staff learning curve, and a higher error rate in the first quarter while habits reset. Then multiply by your charge-out rate, or at least by staff cost. On the hypothetical firm above, avoiding the ledger rise saves £1,320 a year. If moving 40 clients takes even three hours each all-in, that is 120 hours. You can do that arithmetic in your head, and it rarely flatters a rushed move.
What will clients feel?
Bank feeds need re-authorising. Invoice templates change. Logins change. A client who raises sales invoices in the ledger every day feels a migration far more than one who only forwards receipts. Segment before you decide: moving ten clients who never log in is a different project from moving five who live in the software, and the answer may legitimately differ by segment.
Why panic-switching in September is usually worse
Migration costs are paid up front and are certain. Savings arrive monthly and are small. That asymmetry is the whole argument, but there are two others. You carry the filing liability, and transition errors surface at exactly the moment workloads start climbing towards January. And a decision made in irritation tends to optimise for leaving rather than for arriving somewhere better.
A planned review beats a reactive one. Set a window after the January peak. Shortlist against the questions above. Pilot with one or two willing clients whose books are simple and whose VAT quarters align. Measure what the pilot actually cost in hours. Then move segment by segment at quarter boundaries, or conclude, with evidence, that moving is not worth it.
When staying put is the right answer
Staying is the right answer more often than switching articles admit. If clients actively use invoicing, projects, payroll or the app ecosystem, if your team is fluent, and if your whole-stack cost per client is known and defensible within your fees, then staying is a decision rather than inertia. The rise still deserves a response, but the response can be internal: rightsize plans, cut unused add-ons, and look at your own pricing. A vendor increase silently absorbed every year compounds against your margin, so pass it through transparently at renewal or reprice the service, rather than letting it eat the practice quietly.
The honest summary: £2 to £5 per organisation per month is rarely, on its own, a reason to move. Drift across the whole stack, opaque metered charges, or a tool that no longer fits how the practice works are reasons. The September date is simply a good prompt to look.
Where AIONA fits
AIONA is a bookkeeping platform for UK practices, priced per client company at £24 a month with unlimited users, £19 from 26 clients, and AI included with no usage caps or surcharges; your own company is free and so is your first client. If you want to look at a client book without moving anything, AIONA connects to Xero read-only and grades the file for mis-codings, VAT exposure and evidence gaps, and it cannot write anything back. Whatever you decide about your stack, your ledger, documents and audit trail export in full to CSV or Excel at any time, with no exit fee. More at aionatech.com.