Since April 2026, HMRC has required UK sole traders and landlords above an income threshold to keep digital records and submit them quarterly under Making Tax Digital for Income Tax, rather than filing once a year.
For software, this was a manageable change. For practices, the interesting consequence wasn't the filing frequency. It was what happens to a manual process when you run it four times as often.
Check the current thresholds, phase-in dates, and exemptions against HMRC's own guidance before relying on any of this — the rollout has moved before.
Annual processes hide their cost
A once-a-year process absorbs an enormous amount of inefficiency without anyone noticing.
If it takes four rounds of emails to get a client's records, that's mildly annoying once a year. It happens during a known busy period, everyone expects it to be painful, and the pain ends. The inefficiency is real, but it's seasonal, and seasonal pain gets tolerated rather than fixed.
Run the same process quarterly and it stops being a season. It becomes the baseline.
Four times the frequency isn't four times the work
It's usually worse, for two reasons.
The first is that per-cycle overhead doesn't shrink. Working out which clients still owe you records, chasing the ones who haven't replied, checking what came back — that overhead is roughly fixed per cycle, so running four cycles multiplies it directly.
The second is that the gaps between cycles get short enough to overlap. An annual process has eleven months of slack; a quarterly one has weeks. Clients who were slow before are now slow while the next quarter is already starting, and chasing stops being a task with an end.
The filing was never the bottleneck
It's worth being precise about which part of this is automated and which isn't.
The submission is automated. Practice software has handled the mechanics of getting data to HMRC for years, and MTD didn't change that in any way that costs a practice time.
What isn't automated is everything upstream: asking each client for their records, working out whether what arrived is complete and usable, going back for the parts that aren't, and tracking who still hasn't replied. That work is conversational, it's per-client, and it was always the expensive part. Quarterly filing simply removed the eleven months of cover that made it tolerable.
Where the time actually goes
If you break down a quarter's collection work, very little of it is accountancy:
- Identifying which clients owe records this quarter
- Sending each one a request for the specific things they need to send
- Receiving partial, unclear, or wrongly-dated submissions
- Working out precisely what's still outstanding, per client
- Re-asking, then re-asking again
- Checking the complete set against the filing requirement
- Escalating the handful of cases that need a professional judgment
Exactly one item on that list needs a qualified accountant. The rest needs someone reliable with time, which is the most expensive way to buy reliability.
What's worth separating
The useful distinction isn't "manual versus automated". It's judgment versus everything else.
Deciding how an unusual item should be treated is judgment. Noticing that a bank statement is missing eleven days, and asking for the rest, is not. A practice that separates those two cleanly can put the second category somewhere else entirely and keep its qualified time on the first.
That separation is the actual opportunity quarterly filing created. Not faster filing — filing was already fast. A forced, visible accounting of how much unbilled conversational work sits underneath it.