MTD quarterly workflow readiness test
An MTD quarterly workflow is ready only when scope, records, responsibilities, exceptions and approvals can be traced from request to close. Use the 12 tests below on one representative client cycle. Any failed stop condition should be resolved before routine reminders begin.
An MTD quarterly workflow is not ready merely because a reminder template exists. It is ready when the practice can identify the correct client and period, request an approved record set, recognise what comes back, stop on uncertainty and preserve professional approval as a separate decision.
This 12-point test is for UK accountancy and bookkeeping practices preparing an operational record-collection loop. It is not tax, accounting or legal advice. It does not decide whether a client must use Making Tax Digital for Income Tax, which dates apply, which records are sufficient, how transactions should be treated or whether an update should be sent.
How to run the test
Choose one representative client cycle. Use the systems, people and communication routes that would be used in real work, but do not send live reminders or submit anything merely to complete the test.
For each test, record one of three results:
- Pass: the owner, source, action and evidence are clear and can be demonstrated.
- Fix: the process is understood, but a specific control or responsibility is missing.
- Stop: continuing could create a wrong request, missed reply, unsafe data action or unauthorised decision.
The workflow is operationally ready only when every stop has been resolved and each fix has an owner and completion date. A pass here does not establish tax compliance; it establishes that the record-request loop has an inspectable operating boundary.
1. Can you confirm the client scope from a current source?
The test operator should be able to identify the client, relevant business or property activity, period, agreed service and the person who confirmed the scope. The confirmation should point to a current engagement or practice record rather than memory, a copied list or an old reminder campaign.
Stop if the client identity, activity, authority or applicable period is unclear. HMRC’s MTD for Income Tax overview should be checked live by the responsible person when confirming the client’s position.
2. Can you identify one authoritative record list?
Ask the operator where the current required-item list lives. There should be one controlled source, or a documented reconciliation rule when several systems contribute to it.
A spreadsheet, practice-management task and email thread must not silently compete as separate truths. The operator should know which source wins, who can amend it and how a correction reaches the collection workflow.
3. Are the workflow dates visibly sourced?
The cycle should distinguish an official date confirmed for the client from the earlier response date requested by the practice. Record who checked the live source and when.
HMRC’s quarterly update guidance explains the current update periods, deadlines and information sent. Do not copy dates from this article or an old template. Stop if the workflow cannot show the current client-specific instruction used to plan the cycle.
4. Does every responsibility have a named owner?
Identify who owns:
- scope and client responsibility confirmation;
- the required-item list;
- request approval and release;
- reply review and reconciliation;
- professional decisions and corrections;
- exceptions, complaints and access problems; and
- any software write or submission step.
One person may hold several roles, but the decisions must remain distinct. ICAEW’s professional standards guidance discusses service scope, client responsibilities and accuracy. Use the current guidance alongside the practice’s own professional policies.
5. Can the operator produce a minimal request?
Immediately before a request would be sent, the operator should recheck approved sources and remove anything already received, resolved or marked not applicable by an authorised person.
The resulting request should name the period, exact remaining items, approved return route, practice response date and a way to ask questions or report that an item is unavailable. Fail the test if the only available message is a generic request for “everything” or if sensitive information would be invited through an unapproved channel.
6. Does the workflow account for client circumstances?
The process should provide a staffed question route and an alternative for clients who cannot use the default channel. It should be possible to adjust communication without losing the audit trail or creating an unofficial side process.
HMRC’s agent toolkit for preparing clients advises agents to establish roles and responsibilities, set expectations and consider clients’ circumstances and digital capability. A workflow that cannot accommodate an agreed alternative is not ready.
7. Does any reply pause the routine path?
Test what happens when the client replies with a document, a partial answer, a question, “already sent”, “not applicable”, a complaint or an unclear message.
The routine reminder for that request should pause until the reply is classified. A rule that stops only when a particular attachment arrives will miss important responses and can continue chasing a client after the situation has changed.
8. Can received material be reconciled without guessing?
The operator should be able to link a returned item to the client, business, period and request using approved systems. Receipt should not automatically mean sufficient, correctly categorised or ready for the next professional step.
Test duplicate filenames, partial periods and material sent through a different approved route. If the workflow cannot distinguish received, reconciled and professionally reviewed, mark this as a stop.
9. Are exceptions explicit and routable?
Use at least four representative exceptions: delivery failure, disputed request, access problem and unexpected sensitive information. For each one, the workflow should state what stops, what evidence is retained, who is notified and who decides the next action.
An exception queue with no service owner is only a hidden backlog. The test passes when each exception has a named destination and a safe state while it waits.
10. Are professional decisions outside the routine automation?
The collection loop may manage approved requests, factual tracker states and routing. It should not infer a client’s obligation, decide tax treatment, declare records sufficient, approve a correction or authorise a submission.
Ask the operator to show the hand-off to the authorised reviewer and the evidence of that person’s decision. Stop if a tracker state, model output or successful upload can bypass professional approval.
11. Are writes and failures verifiable?
Where the workflow updates a tracker, files material or interacts with software, test the response to a timeout, permission failure and ambiguous result. It should not retry indefinitely or convert uncertainty into success.
The workflow should preserve the attempted action, result, time and next owner. A final status should be based on a verified response or a human decision, not the absence of an error message.
12. Can the cycle close with evidence?
At close, the practice should be able to show:
- the scope and dates used;
- the approved item list;
- requests and material replies;
- reconciled states and unresolved exceptions;
- professional approvals and corrections;
- any permitted software action and its result; and
- the person and time that authorised closure.
The cycle is not complete because reminders stopped. It is complete when the responsible person can explain the final state from the retained evidence.
Readiness decision
Use a simple rule after the test:
- Ready for a bounded live pilot: all 12 tests pass, or remaining fixes are non-safety improvements with named owners.
- Repair before piloting: one or more fixes affect ownership, reconciliation, accessibility or evidence, but the workflow can remain safely paused.
- Not ready: any stop remains around identity, scope, dates, permissions, replies, sensitive information, professional judgement or ambiguous writes.
If the process is not ready, use the MTD quarterly record-collection checklist to rebuild the cycle. For a wider operational diagnostic, the AI chase map helps map hand-offs and residual human work.
Sources and review status
Author: Babagana Zannah. Published 13 August 2026 and last updated 13 August 2026. Sources checked 13 August 2026. Next editorial review due 13 November 2026.
The following sources were re-opened on the source-check date. Recheck them sooner when HMRC guidance, professional standards or the practice’s operating scope changes.
- HMRC: Use MTD for Income Tax, before you use this guide
- HMRC: Use MTD for Income Tax, send quarterly updates
- HMRC: Get ready for MTD, preparing your clients
- ICAEW: Professional standards for MTD for Income Tax services
Questions owners ask
What does MTD quarterly workflow readiness mean?
It means the practice can run a defined record-collection cycle with current client scope, named owners, controlled requests, visible exceptions and separate professional approval. It does not determine a client's tax obligations or submission readiness.
Is this readiness test an MTD compliance checklist?
No. It is an operational test for a practice workflow. Use current HMRC guidance and an appropriately authorised professional to determine obligations, dates, records, treatment and submission decisions for each client.
Should a practice automate a workflow that fails the test?
Not yet. Automation can repeat an unclear or unsafe process faster. Resolve source ownership, stop rules, reply handling, permissions and approval boundaries first, then test a bounded workflow with representative cases.
How often should an MTD record-collection workflow be retested?
Retest after a material guidance, software, engagement, responsibility or data-flow change, and before relying on a redesigned workflow. Practices should set a review frequency that reflects their own risk and change environment.