Accounting services target audience: work and obligations

Define an accounting services target audience by the work it needs completed, not by a vague label such as "small businesses". A useful definition combines entity, activity, obligation, record condition, decision-maker and delivery constraints.

What to take away

Start with one service decision

State what the research must decide. You might be testing a bookkeeping package for new employers, an MTD readiness service for landlords or directory filters for limited companies seeking annual accounts support.

List assumptions as questions. Does the customer keep digital records? Who approves figures? Is the deadline known? Which adviser already acts? What makes the current process difficult? Do not recruit participants until the team knows what evidence would change the offer.

GOV.UK's guidance on planning user research recommends agreeing objectives, identifying groups and selecting methods that answer the priority questions. It was written for government services, so adapt the method rather than presenting it as an accountancy rule.

Segment by work, not stereotype

Record variables that affect delivery:

Avoid invented attitudes such as "traditional shop owner" or "tech-savvy founder". Observe behaviour and quote it only with consent.

Recruit actual or likely users

The Government Digital Service says effective research participants should be actual or likely users and advises inclusion of people who need support or use assistive technology. Build a screening form from the segment variables and conceal any preferred answer.

Recruit through several routes. Existing clients can explain current delivery but may overrepresent people who already trust the practice. Include recent switchers, self-servers, people who abandoned a search and users of competing routes. Do not let one referral partner supply the whole sample.

Connect stated need with evidence

Ask participants to walk through a recent accounting task using redacted records where safe. Capture what triggered action, which information was missing, who decided, how long each stage took and what outcome was accepted.

Separate a requested feature from the underlying job. "I want an app" might mean faster document capture, visible progress or fewer repeated questions. Test those needs independently.

Write an evidence-based audience statement

Use a form such as: "England-based owner-directors of single-entity service companies who keep records in supported software, need defined annual accounts and Corporation Tax work, and can supply approved records by the agreed cut-off."

For contrast, consider an England-based individual landlord who wants to check whether HMRC's MTD for Income Tax requirements apply, has incomplete digital rental records, makes decisions about the records personally and needs help organising them remotely. That segment points to a readiness and records-support service; landlords whose accountant already manages those tasks may need a different offer.

Add exclusions and referral conditions. State the sample, recruitment source and unanswered questions. Review the definition when legislation, software, capacity or observed customer behaviour changes. A narrow audience that the team can serve well is more useful than a large audience assembled from assumptions.

Check whether the audience is reachable

Map how matching people currently seek help: an existing adviser, professional directory, local network, software marketplace, search engine or referral. For directory recruitment, potential routes include ICAEW's Find a Chartered Accountant directory, ACCA's Find an Accountant directory and AAT's Find a Bookkeeper directory. Record the evidence for each channel and any gatekeeper. A group may have a genuine need but be uneconomic to reach through the proposed route.

Run a small recruitment and qualification pilot. Count invitations, eligible responses, completed sessions, service-ready cases and exclusions separately. Compare acquisition effort with delivery capacity. If recruitment reaches only confident digital users, add an assisted route before claiming to understand the whole segment.

Keep research participation separate from marketing consent. If research involves personal data, identify the appropriate lawful basis under UK GDPR and follow ICO guidance; participation consent and marketing consent are not interchangeable. Publish only aggregated or safely anonymised findings, and have the privacy owner check any case example before release.

Before you act

Common questions

What should a useful audience definition combine?

It combines entity, activity, obligation, record condition, decision-maker and delivery constraints. The article says to choose an audience through the work it needs completed rather than a vague label such as small businesses, and to record variables that affect delivery.

Why should recruiters use several routes to find participants?

Existing clients can explain current delivery but may overrepresent people who already trust the practice. The article advises including recent switchers, self-servers, people who abandoned a search and users of competing routes, and not letting one referral partner supply the whole sample.

How should a team handle a requested feature such as an app?

Separate the requested feature from the underlying job. The article says a request for an app might mean faster document capture, visible progress or fewer repeated questions, and those needs should be tested independently rather than accepted at face value.