MTD for Income Tax: the capacity problem nobody is costing

The dates, so they are in one place

Qualifying income is turnover from self-employment and property before expenses, not profit, and it is measured two years back. So the population that joins in April 2026 was set by their 2024/25 figures — which means you already have the data to work out exactly who is affected. You are not waiting on anything.

The quarterly update deadlines are fixed and they do not move: 7 August, 7 November, 7 February and 7 May, with the final declaration on the following 31 January. For the first cohort that means the first update is due 7 August 2026 and the first final declaration on 31 January 2028.

HMRC has said it will not issue penalty points for late quarterly updates during 2026/27. Read that carefully: it is a one-year grace for the first cohort only, it does not cover the final declaration, and from 2027/28 the points system applies in full — a point per missed deadline, £200 at four points, and a further £200 for each miss after that. The grace year is time to get the process right, not permission to be late.

The arithmetic nobody is doing

Take a practice with 400 clients, of whom 180 are sole traders or landlords over the threshold. Today those 180 generate 180 filing events a year. From April 2026 they generate 900 — four quarterly updates each, plus a final declaration.

The total work is not five times larger. The same income and expenditure still has to be recorded once. What changes is the number of occasions on which somebody has to make a client produce records, check what arrived, and file something by a date. That number goes up five-fold, and it is the part of the job that has never scaled well, because it is mostly chasing.

Those minutes are an assumption, not a finding, and the honest answer is that nobody knows yet because the regime has not run. Replace them with your own number. But run the exercise, because the gap between fifteen minutes and an hour is the difference between absorbing this and hiring for it, and the thing that decides which end you land on is not skill — it is whether the records turn up without being asked.

February is the problem

Look at the calendar rather than the list. The 7 February quarterly update lands seven days after the 31 January self assessment deadline.

Every practice in the country runs January at maximum. Most run it on goodwill, late evenings and a team that is counting down. On 1 February that team has historically got a fortnight of recovery, catch-up and holiday. From 2027 they will have a week, and then a deadline for a few hundred clients whose records they have not started chasing, because they were doing January.

This is not a compliance risk, it is an attrition risk. The profession is already losing people to workload — and February is where a practice that has not planned for this will find out what its team will and will not tolerate. If you do one piece of planning, do this one: decide now who is chasing Q3 records, and when they start. The answer has to be December, which means it collides with something else, which is the whole point.

Why the software question is a distraction

The software question is close to settled. Every serious practice package files quarterly updates; you will pick one, it will work, and it will not differentiate you.

What software cannot do is make a client send you their records. Five times a year, for several hundred clients, somebody has to. If your current process for getting records is an email in May and three reminders in December, that process is about to run five times and it will break the first time it does.

The practices that will cope are the ones where records arrive because a system pulls them — bank feeds connected, receipt capture in the client's hands, a standing date rather than an annual ask — rather than because somebody remembered to chase. That is a workflow project, it takes a quarter or two to embed, and it is the reason the useful work has to happen before April rather than after it.

What to do with the three quarters you have left

This is a sequence, and the order matters more than the pace.

Now: find out how big it is

Run the list. You hold 2024/25 figures for every client, so you can produce an exact count of who crosses £50,000 in an afternoon. Then do the same for £30,000 and £20,000, because that tells you the shape of 2027 and 2028 and stops you solving this three times. Most owners are surprised in both directions — the April 2026 number is smaller than feared, and the April 2028 number is much larger than anyone has looked at.

Next: fix how records arrive, for one segment

Do not attempt all of them. Take the largest coherent group — usually landlords, because their records are the most uniform — and get that group onto connected bank feeds and a standing monthly rhythm before Christmas. One segment genuinely working beats five segments half-migrated, and it gives you a real per-client time figure to plan the rest with.

Then: price it, before you deliver it

Five filing events cannot sit inside a fee built for one. Decide the number now, while it is a calm conversation about a known change, rather than in August when you are doing the work and resentful about it. Firms that reprice in advance report a change in fee; firms that reprice afterwards report a dispute.

Last: decide who owns it

Name one person who owns MTD delivery, with the authority to set the record deadlines and hold clients to them. Unowned work defaults to whoever cares most, and in a founder-led practice that is the owner. A regime with twenty filing dates a year cannot be run out of the owner's inbox.

The honest framing

MTD is not an opportunity and it is not a catastrophe. It is a permanent five-fold increase in the number of times you have to make something happen on a date, landing on a profession that is already turning work away for want of people.

Practices that treat it as a software purchase will absorb it with overtime in year one and lose somebody in year two. Practices that treat it as a workflow problem — how records arrive, who owns the chase, what it costs, what gets standardised — will come out the other side with a better-run business that happens to also be MTD-compliant.

The second group is not working harder. They started three quarters earlier, on the boring part.