Segmenting a media agency's target audience by the job, not the logo

A media buying agency's target audience is not "all businesses that advertise". It is a set of buyers with a shared media decision, enough budget or complexity to seek help, and a procurement route the agency can serve. Define that group before choosing research participants or writing a proposition.

What to take away

Segment by the job, not the logo

Start with the trigger. One advertiser may need national television planning; another may need local paid search; a third may want independent scrutiny of programmatic supply. Company size can affect the solution, but it does not reveal the decision on its own.

Write one sentence that names the buyer, problem and boundary. For example: an England-based marketing team that buys several digital channels but lacks a common reporting definition. This is a research hypothesis, not a claim about market prevalence.

Use real buying conditions

Separate prospects by variables that change delivery:

The IPA guide to finding an agency asks clients to consider why they are reviewing, how procurement participates and which pitch model fits their way of working. Those questions provide better segmentation clues than broad industry labels.

Agency type changes the buying conditions too. A regional independent in an English city such as Manchester or Leeds usually meets owner-managed advertisers with smaller budgets and quicker sign-off. A London network agency tends to pitch larger national brands.

Public sector prospects often buy media through a Crown Commercial Service framework, which adds a tender timetable and its own pricing rules.

Recruit people who make and use the decision

Include the budget sponsor, day-to-day marketing owner and specialists who will rely on agency outputs. Finance, procurement, legal, data protection or brand teams may become essential participants when the contract or campaign risk warrants it.

Government Digital Service guidance on finding user-research participants recommends recruiting actual or likely users and considering groups that may otherwise be excluded. It describes UK government-service practice, so a commercial agency should adapt the principle and document its own sample.

Set a rough recruitment target rather than an open-ended list. For one segment, plan 8 to 12 interviews, with at least three in each distinct buying role. The right number depends on how many roles the segment contains and on when answers stop changing, so stop once two consecutive interviews repeat the same buying conditions.

Add audience evidence cautiously

Ofcom's Adults' media use and attitudes research covers UK adults aged 16 and over. It can inform the people an advertiser may want to reach, but it does not identify the agency's business buyers.

Secondary data can help frame recruitment. It cannot reveal undocumented approval politics, reporting frustrations or internal capability. Interviews are needed for those.

Decide whether the segment is serviceable

After research, test four conditions. Can the agency reach these buyers? Does the problem occur often enough to sustain a service? Can the team solve it with evidence? Will the buyer's procurement and risk requirements fit the operating model?

Apply the same tests to the retailer segment recorded below. Reach is practical where the agency already buys local radio. Frequency holds while store openings continue. Evidence exists if the agency can show comparable retail campaigns. Procurement fit is simple when one marketing manager signs.

Keep competing segments separate until those answers are known. A local direct-response buyer and a multinational brand team may both purchase media, but they can require different people, contracts and tools. The final target definition should state who is excluded and why, then remain on hold until the sample and interpretation receive named editorial review.

Record the result so another planner can use it

A usable segment note names the decision maker, buying event, channels, geography, internal capability and disqualifying conditions. Attach the interview dates and source links, then mark assumptions separately from observed evidence. Avoid adding a demographic detail simply because a persona template contains a box for it.

An illustrative note, not a real client, reads like this:

Revisit the note when account ownership, regulation or the advertiser's channel mix changes. Audience definition is an operating input, not a slogan. Its quality becomes visible when a colleague can recruit the right respondent or qualify an enquiry without asking what the segment description really means.

Before you act

Common questions

How should an agency start defining its target audience?

Start with the trigger. Write one sentence naming the buyer, problem and boundary, and treat it as a research hypothesis rather than a claim about market prevalence. Company size can affect the solution but does not reveal the decision on its own.

Which variables should separate prospects?

Separate prospects by variables that change delivery: channels and territories in scope, whether planning or activation stays in-house, regulated products, ownership of audience data and platform accounts, procurement requirements, and the frequency and value of buying decisions. The IPA guide's questions about review, procurement and pitch model offer better clues than industry labels.

What should a usable segment note contain?

A usable segment note names the decision maker, buying event, channels, geography, internal capability and disqualifying conditions. Attach interview dates and source links, and mark assumptions separately from observed evidence. Revisit the note when account ownership, regulation or the advertiser's channel mix changes, because audience definition is an operating input.