Most buyer personas are fiction.
They are created in a two-hour workshop where the VP of Sales describes their favourite customer and the marketing team turns it into a slide with a stock photo, a name like "Strategic Sarah," and a list of generic pain points. The persona goes into a shared drive. Nobody opens it again. Six months later, a new PMM joins and creates another one from scratch.
This is not research. It is creative writing.
The personas that improve GTM come from a different process. They are discovered through structured research, validated with data, and updated when the market shifts. They answer specific questions that change how you sell, what you build, and where you spend money.
This framework shows you the 3-Stage Persona Discovery Process: how to move from assumptions to evidence to a living document your team uses every week.
What is a Buyer Persona?
A buyer persona is a research-based profile of a specific buyer type within your ICP. It describes who they are, what they care about, how they buy, and what language they use to describe their problems. A good persona changes how you write copy, build demos, train sales reps, and prioritise features. A bad persona sits in a slide deck and does nothing.
Why Personas Fail: The 3 Failure Modes
Before building a better process, understand why the existing one breaks.
Failure Mode 1: Invented in Workshops
A cross-functional team gathers in a room. They brainstorm who the buyer is based on collective memory. The loudest voice wins. The persona reflects internal assumptions, not external reality.
The problem: workshop personas reflect what the team wants the buyer to be, not who the buyer actually is. They over-index on the deals Sales remembers and ignore the deals that never reached Sales in the first place.
Failure Mode 2: Based on One Loud Customer
A major customer gives detailed feedback. The team builds the persona around this one account. Every feature request, every complaint, every preference comes from a single source.
The problem: one customer is not a pattern. Their needs might be unique to their size, industry, or internal politics. Building a persona around one voice means optimising for an audience of one.
Failure Mode 3: Never Updated
The persona was accurate two years ago. Since then, the product has changed, the market has shifted, and three new competitors have entered. But the persona still says the buyer cares about "ease of use" and "time savings" because nobody has gone back to check.
The problem: stale personas create stale messaging. Teams keep solving yesterday's problems while buyers have moved on to new ones.
"The most dangerous persona is the one everyone agrees on but nobody has validated. Agreement without evidence is groupthink."
The 3-Stage Persona Discovery Process
This process moves from internal assumptions to external evidence to quantitative validation. Each stage builds on the previous one. Do not skip stages.
Stage 1: Internal Alignment Using the Empathy Map
Duration: 1-2 days
Who is involved: PMM, Sales lead, CS lead, Product lead
Purpose: Capture what the team currently believes about the buyer. This is not the persona. It is the hypothesis.
Use an empathy map to structure the assumptions. Capture these seven dimensions for each buyer type:
| Dimension | What to Capture | Example (Head of Talent) |
|---|---|---|
| Demographics | Age range, education, career stage | 35-45, degree-educated, 10+ years in HR/People |
| Firmographics | Company size, industry, stage | 200-500 employees, B2B SaaS, Series B-C |
| Psychographics | Values, beliefs, identity | Believes hiring is broken, values data over gut feel |
| Jobs to Be Done | Functional, emotional, social jobs | Fill roles fast, look competent to the board, build a team others envy |
| Gains | What success looks like | Reduced time-to-hire, higher offer acceptance, consistent quality |
| Pains | Frustrations, obstacles, fears | Too many open roles, hiring managers who ghost, high agency spend |
| Triggers | Events that start a buying process | New funding round, missed quarterly hiring targets, exec mandate |
Document every assumption. Mark each with a confidence level: high (based on data), medium (based on repeated anecdotes), or low (based on one person's opinion). The low-confidence items become your interview priorities.
Stage 2: Qualitative Interviews (10-15 Conversations)
Duration: 2-3 weeks
Who to interview: 10-15 people who match your assumed persona. Mix of current customers (5-7), recently lost prospects (3-4), and people in your ICP who have never heard of you (2-3).
Purpose: Validate or disprove your Stage 1 assumptions with real buyer language and real buyer priorities.
Every interview must answer three questions in plain language:
- Who are you? What is your role, your team, your reporting line, your daily reality? What does success look like in your job this year?
- What do you do? Walk me through how you handle [the problem your product solves] today. What tools do you use? What is manual? What breaks?
- Why do you do it that way? What have you tried before? Why did you switch or stay? What would need to change for you to try something new?
These three questions surface the buyer's context, behaviour, and decision triggers. Everything else is detail.
Interview tips that matter:
- Record every call. Transcribe it. Your memory is unreliable.
- Ask "tell me about the last time..." instead of "do you usually...?" Past behaviour beats stated preference.
- When they say something interesting, ask "why?" at least twice. The first answer is surface-level. The second reveals motivation.
- Do not pitch your product. You are researching, not selling. The moment you start explaining features, the interview is contaminated.
- 15 interviews is not a magic number. Stop when you hear the same themes three times in a row. That is saturation.
Stage 3: Quantitative Survey Validation
Duration: 2-3 weeks (design, distribute, analyse)
Sample size: 50-200 respondents
Purpose: Confirm that the patterns from 15 interviews hold across a larger population. Quantify the priorities.
Do not run a survey before doing interviews. Surveys confirm patterns. They do not discover them. If you survey before interviewing, you will ask the wrong questions and get confident answers to irrelevant topics.
What to confirm in the survey:
- Priority ranking of the top 5 pains you heard in interviews. "Rank these challenges from most to least pressing."
- Current solution landscape. "Which of these tools do you currently use for [problem]?"
- Buying triggers. "Which of these events would prompt you to evaluate a new solution?"
- Decision criteria. "When evaluating a new tool, which factors matter most?" (Force-rank, not checkbox.)
Survey results give you percentages. "73% of Heads of Talent say time-to-hire is their top pain point" is more persuasive in a board meeting than "we heard it in interviews."
The 3 Questions Every Persona Must Answer
Strip away the formatting, the stock photos, and the clever names. A persona is useful if and only if it answers three questions in plain language:
- Who are you? Role, seniority, team size, company profile, reporting line. Enough detail that a new sales rep can picture the person on the other side of a Zoom call.
- What do you care about? Top 3 pains, primary goals this quarter, what they are measured on, what keeps them up at night. Not generic concerns. Specific, ranked priorities.
- How do you buy? Who else is involved in the decision? What is the typical budget range? Where do they research solutions? What triggers a buying process? What kills a deal?
If your persona cannot answer these three questions with specifics, it is not finished.
Filled Persona Template: Head of Talent at a Mid-Market SaaS Company
This is what a complete, research-backed persona looks like. Every field is filled with specifics. No placeholders.
| Field | Detail |
|---|---|
| Job Title | Head of Talent / Director of Talent Acquisition |
| Reports To | VP of People or Chief People Officer |
| Team Size | 3-6 recruiters, 1 coordinator, sometimes 1 employer branding specialist |
| Company Profile | B2B SaaS, 200-500 employees, Series B or C, growing 30-50% YoY, 40-80 open roles at any time |
| Experience | 8-15 years in talent/recruitment, moved from agency to in-house 4-6 years ago |
| Top 3 Pains | 1. Time-to-hire is too long (averaging 52 days, target is 35). 2. Quality of hire is inconsistent across hiring managers. 3. Agency spend is eating the budget (£180K last year, up 40% from previous year). |
| Primary Goals This Quarter | Reduce time-to-hire to 40 days. Cut agency dependency by 30%. Improve hiring manager satisfaction score from 6.2 to 7.5. |
| What They Are Measured On | Time-to-hire, cost-per-hire, offer acceptance rate, hiring manager satisfaction, diversity of pipeline |
| Current Tools | Greenhouse (ATS), LinkedIn Recruiter, a spreadsheet for pipeline tracking, Slack for hiring manager comms |
| Frustrations with Current Setup | Greenhouse does not surface hiring quality data. LinkedIn Recruiter is expensive and response rates are dropping. No single view of where each role stands. Hiring managers do not give timely feedback on candidates. |
| Buying Triggers | New funding round with aggressive headcount targets. CEO or CPO mandates faster hiring. A bad hire in a senior role that triggers a process review. |
| Decision Process | Head of Talent evaluates 3-4 options. Shortlists 2 for demo with VP People and 1 hiring manager. VP People has final budget sign-off. IT does a security review. Process takes 4-8 weeks. |
| Preferred Channels | LinkedIn (daily), People-focused Slack communities (People Collective, SHRU), industry newsletters (Hung Lee's Recruiting Brainfood), peer recommendations at events |
| Language They Use | "Quality of hire" not "talent quality." "Hiring velocity" not "recruitment speed." "Pipeline health" not "candidate funnel." "Structured interview process" not "assessment framework." |
| What Kills a Deal | Clunky UX that hiring managers will refuse to use. No ATS integration with Greenhouse. Pricing that requires annual commitment with no pilot option. Sales rep who cannot answer technical questions about data handling. |
Notice the specifics. Not "they care about efficiency." Instead: "time-to-hire is averaging 52 days, target is 35." Not "they use recruiting tools." Instead: "Greenhouse for ATS, LinkedIn Recruiter, a spreadsheet for pipeline tracking." This level of detail is what makes a persona useful.
The Persona Document: What to Include and What to Skip
Keep the document to two pages maximum. Include:
- The filled template above (one page).
- Three direct quotes from interviews that capture their voice.
- A "day in the life" paragraph. Not a narrative essay. A bullet list of what Monday morning looks like for this person.
- Messaging implications: what to lead with, what to avoid, which proof points matter.
Skip:
- Stock photos or AI-generated images. They add nothing and make the document look like a marketing class assignment.
- Names like "Strategic Sarah" or "Data-Driven Dave." Use the job title. It is clearer and more professional.
- Personal hobbies. Unless they directly affect buying behaviour, nobody cares that this persona "enjoys hiking on weekends."
- Demographic details beyond what affects the sale. Age, gender, and education matter only if they change how you sell. Usually they do not.
Common Persona Traps
Trap 1: Too Many Personas
A startup with 50 customers creates 7 personas. Each one is thin. None has enough research behind it. The team is paralysed trying to serve everyone.
Start with 2-3 personas maximum. The primary buyer (who signs the contract), the primary user (who uses the product daily), and one secondary influencer (who has veto power). Add more only when you have the research and the resources to serve them differently.
Trap 2: Confusing Persona with ICP
ICP describes the company you sell to (size, industry, stage, signals). Persona describes the person within that company. You need both. ICP tells you where to fish. Persona tells you what bait to use.
Trap 3: Building Personas but Not Using Them
The persona exists. Nobody references it. Content is written without checking persona language. Sales decks do not reflect persona priorities. Product roadmaps do not link features to persona pains.
Fix this by embedding personas into workflows. Every content brief should reference the target persona. Every sales enablement session should start with "who are we talking to and what do they care about?" Every product requirements document should link to a specific persona pain.
How to Keep Personas Alive
A persona is not a one-time deliverable. It is a living document. Here is the maintenance schedule:
- Monthly: Review win/loss notes and customer feedback for signals that contradict or confirm the persona. Update the "top pains" and "buying triggers" sections if patterns shift.
- Quarterly: Run 3-5 fresh interviews with people who match the persona. Check whether the priorities, language, and buying process still hold.
- Annually: Full persona refresh. Re-run the 3-stage process. Markets change. Competitors emerge. New tools shift expectations. A persona from 2024 is not automatically valid in 2026.
Assign one person as the persona owner. In most companies, this is the PMM. The owner does not do all the research alone, but they are responsible for keeping the document current and making sure the team uses it.
From Persona to Action: What Changes
A completed persona should change at least four things in your GTM:
- Content topics. Write about what the persona searches for, not what your product does. The Head of Talent persona above is searching for "how to reduce time-to-hire" and "structured interview best practices," not "AI-powered recruitment platform features."
- Sales messaging. Lead with the persona's top pain, not your product's top feature. For the Head of Talent: "Your time-to-hire is 52 days. We get mid-market SaaS companies to 35 in the first quarter."
- Demo flow. Show the features that solve the persona's top 3 pains in the order they rank them. Do not show every feature. Show the three that matter.
- Product roadmap input. Link feature requests to persona pains. "This feature addresses the number 2 pain point for our primary persona" is more compelling than "12 customers asked for this."
Frequently Asked Questions
How many personas should we have?
Start with 2-3. One primary buyer, one primary user, and one influencer or secondary buyer. Do not create more personas than you can properly research and actively use. Five thin personas are worse than two deep ones. Add new personas only when a distinct buying pattern emerges that your existing personas do not cover.
How often should we update them?
Monthly: check signals from win/loss and customer feedback. Quarterly: run 3-5 fresh interviews. Annually: full refresh of the 3-stage process. If your market is changing fast (new competitors, new regulations, major platform shifts), accelerate the cycle. If your market is stable, the quarterly check is enough.
What if internal teams disagree on who the buyer is?
This is common and healthy. Sales often sees a different buyer than Product because they interact at different stages. Do not resolve the disagreement in a meeting. Resolve it with data. Run Stage 1 to capture each team's assumptions explicitly. Then run Stage 2 interviews to test both hypotheses. The interviews will reveal who is right. If both buyers exist, you may need separate personas for each. If one team is wrong, the interview evidence makes the conversation factual instead of political.
How do we know when a persona is good enough to use?
A persona is good enough when it passes three tests. First: a new sales rep can read it and accurately describe who they are selling to and what that person cares about. Second: a content writer can use it to generate 10 blog topics the persona would search for. Third: the product team can point to specific features that address the persona's top 3 pains. If any of these tests fail, the persona needs more detail or more research behind it.