Audience research
Part of Seven steps to digital PR audience research that earns coverage
Four data sources for building digital PR buyer personas
Most teams guess their audience. This article explains how to build digital PR buyer personas from four verifiable data sources and a repeatable six-step process.
What to take away
- The most common mistake is inventing personas from team opinion rather than published data.
- Four sources carry most of the weightONS income data, Pew generational research, BSI AI guidance and your own first-party records.
- A persona needs a job role, a budget holder, a trigger event and a measurable outreach signal. Without those four, it is a sketch.
- Build personas in a numbered sequence, then test them against reply and coverage rates before you commit budget.
- Personas inform the audience research guide for England, which sets out how the wider programme fits together.
Why personas fail
Personas usually fail because they describe a demographic, not a decision. A persona that says "marketing manager, 35, likes podcasts" tells an outreach team nothing about who signs the budget or what prompts a search for help.
Weak vs useful persona
Weak persona
- Focus
- Demographic
- Key question
- Who are they?
- Trigger
- Likes podcasts
- Proof
- None
- Source
- Old pitch decks
Useful persona
- Focus
- Decision
- Key question
- Who owns problem?
- Trigger
- Event that makes them act
- Proof
- Proof needed to reply
- Source
- Published data
A useful digital PR buyer persona answers three questions. Who owns the problem? What event makes them act? What proof do they need before they reply to a pitch? Everything else is decoration.
The second failure is source quality. Teams recycle assumptions from old pitch decks. Published data is slower to gather but survives client scrutiny, which matters when a retainer is reviewed.
Four data sources
ONS income and spending
Household income shapes what a buyer can approve. The ONS publishes disposable income estimates by household type, which lets you split prospects by purchasing power rather than by job title alone.
Four data sources
Source
- ONS income
- Purchasing power bands
- Pew research
- Cohort news habits
- BSI guidance
- AI governance reference
- Reply data
- Confirm patterns
What it gives you
- ONS income
- Pew research
- BSI guidance
- Reply data
Use this to set realistic retainer bands. A persona attached to a business with limited disposable income will not approve a five-figure programme, however good the pitch.
Pew generational research
The Pew Research Centre maintains a public hub of age and generational research covering how different cohorts consume news and evaluate information. That evidence helps you decide which publications and formats each persona actually reads.
Generational labels are blunt instruments, so treat them as a starting filter. Confirm the pattern with your own reply data before you build a whole strategy on it.
BSI guidance on AI
Buyers increasingly ask how AI is used in content production and outreach. The British Standards Institution publishes guidance on AI and digital trust, which gives a recognised reference point when a prospect raises governance questions.
A persona that includes a compliance concern needs a documented answer. Citing a standards body is stronger than asserting your own good practice.
Your own first-party records
CRM notes, pitch replies and coverage logs are the only source that describes your actual buyers. Count how many replies came from each role, which trigger events preceded them and how long each deal took. Small samples still beat no data, provided you label them as internal figures.
A six-step build
- List the job titles that have replied to outreach in the past year.
- Group them by budget authority, not seniority alone.
- Attach a trigger event to each group, such as a funding round or a new product launch.
- Cross-check each group against ONS income bands and Pew cohort data.
- Write one page per persona covering role, trigger, proof needed and preferred publication.
- Review the set every quarter against reply and coverage rates.
Step four is where most teams skip work. Cross-checking takes an afternoon and prevents a persona set that only reflects internal assumptions.
Signals to track
Once personas exist, tie each one to a signal you can observe. For a head of content, that may be a rise in branded search. For a founder, it may be a competitor mention in a target publication. The companion piece on digital PR target audience data and sources covers how to collect and rank those signals.
Track three numbers per persona: reply rate, coverage rate and average deal value. If a persona generates replies but no coverage, the media list is wrong. If it generates coverage but no revenue, the budget authority is wrong.
For example, a team paying £400 a month for outreach might find that founder-led personas reply twice as often as marketing-manager personas but convert at half the rate. That pattern is worth knowing before the next pitch cycle.
Common questions
How many personas should a digital PR team build?
Three to five is usually enough for an agency serving small and mid-sized clients. More than five fragments your media list and makes reporting harder.
Can personas be built without paid data tools?
Yes, if you combine free published sources with your own CRM records. The ONS and Pew datasets are free, and internal reply logs cost nothing but time to compile.
How often should personas be updated?
Quarterly is sensible for most teams, or sooner if a major client sector shifts. Review reply and coverage rates at the same time so the update is evidence-led.
Do personas replace audience research?
No. Personas are one output of audience research. They sit alongside market sizing, message testing and media mapping, not in place of them.
