User persona: how to create customer profiles that actually work

90% of the user personas I've seen in my career are fiction. "Maria, 34 years old, likes yoga and shops online." That's not a persona — it's a useless stereotype that doesn't help make any design, copy, or conversion decision.

A well-crafted user persona is a decision-making tool. When your team debates whether the CTA should say "Start free" or "Request a demo," the right persona gives you the answer.

What a user persona really is

A user persona is a semi-fictional representation of your ideal customer based on real research data. Not on assumptions, not on internal brainstorming, not on what "we think" the customer wants.

Good personas answer three questions:

What do they want to achieve? (Jobs to be done)
What's stopping them? (Barriers and friction)
How do they make the purchase decision? (Process and criteria)

Everything else — age, hobbies, stock photos — is context. Useful, but secondary.

Why user personas fail

Mistake 1: Based on assumptions

The marketing team meets, brainstorms, and "creates" a persona based on what they think their customer is. Without interviews, without analytics data, without surveys. Result: a persona that reflects the team's beliefs, not market reality.

Mistake 2: Too demographic

"Woman, 28-35 years old, urban, medium-high income." That describes millions of people who make completely different decisions. Demographics don't predict behavior. What predicts behavior are motivations, frustrations, and usage context.

Mistake 3: Created and forgotten

A persona in a nice PDF that nobody consults is a waste. Personas should be present in every design, copy, pricing, and feature decision.

Mistake 4: Too many personas

If you have 8 personas, you have none. It's impossible to optimize the experience for 8 different profiles. Maximum 3-4 personas, prioritized by business value.

How to create data-driven user personas

Step 1: Collect quantitative data

Before talking to anyone, look at what you already know:

Google Analytics: demographic data, devices, traffic sources, most visited pages, behavior flow
CRM: real customer data (company size, industry, average ticket, sales cycle)
Heatmaps and recordings: how users actually navigate your website (Hotjar, Microsoft Clarity)
Post-purchase surveys: why they bought, what almost made them leave

Step 2: Qualitative interviews

Talk to 8-12 real customers (and 3-4 who didn't buy). Key questions:

"What problem were you trying to solve when you found us?"
"What alternatives did you consider?"
"What almost stopped you from buying?"
"How would you describe what we do to a colleague?"
"What do you value most about our product/service?"

Record the interviews (with permission) and look for patterns in the responses.

Step 3: Identify patterns and segment

Group responses by behavioral patterns, not demographics. Look for:

Similar motivations: what do they want to achieve?
Common frustrations: what's stopping them?
Similar decision processes: how do they buy?
Recurring objections: what holds them back?

Step 4: Create the persona

Use this template:

User persona template

NAME: [Representative name]
ROLE: [Position / Role in the purchase decision]

JOBS TO BE DONE (what they want to achieve):
- Primary job: ___
- Secondary job: ___
- Emotional job: ___

FRUSTRATIONS (what's stopping them):
- Frustration 1: ___
- Frustration 2: ___
- Frustration 3: ___

DECISION CRITERIA (how they choose):
1. ___ (most important)
2. ___
3. ___
4. ___ (least important)

COMMON OBJECTIONS:
- "___"
- "___"

CHANNELS (where they seek information):
- ___
- ___

REAL QUOTE (verbatim from interview):
"___"

SUCCESS METRIC (how they measure if the solution works):
- ___

Complete example

NAME: Carlos, the pressured CMO
ROLE: CMO / Marketing Director at an ecommerce (10-50 employees)

JOBS TO BE DONE:
- Primary: Increase the ecommerce conversion rate without increasing ad spend
- Secondary: Demonstrate ROI to the CEO with concrete data
- Emotional: Feel in control of growth, not dependent on luck

FRUSTRATIONS:
- Invests heavily in traffic but conversion doesn't improve
- Previous agencies delivered nice reports with no real impact
- Internal team lacks CRO/experimentation expertise

DECISION CRITERIA:
1. Demonstrable results (case studies with metrics)
2. Clear and transparent methodology
3. Specialization (doesn't want a generalist agency)
4. Price (important but not decisive if ROI is clear)

OBJECTIONS:
- "We already tried an agency and it didn't work"
- "How do I know the results are real and not just luck?"

CHANNELS:
- LinkedIn, specialized blogs, marketing podcasts
- References from other CMOs

REAL QUOTE:
"I need someone who obsesses over my conversion as much as I do"

SUCCESS METRIC:
- Measurable conversion rate increase within 3 months

How to use personas to improve conversion

In your website copy

Every headline, every CTA, every bullet point should respond to a job, a frustration, or an objection of your primary persona.

Job: "Increase your conversion without increasing your ad spend" (responds to Carlos's primary job)
Frustration: "No more nice reports that don't accomplish anything" (addresses the frustration with previous agencies)
Objection: "Real cases with verifiable metrics" (dismantles the objection that "it didn't work before")

In experience design

If your primary persona looks for "demonstrable results," your case studies can't be hidden on a secondary page. They should be in the hero, above the fold, on every service page.

In content strategy

Create content that answers the questions your persona asks at each funnel stage:

Awareness: "Why isn't my ecommerce converting?" (educational blog post)
Consideration: "CRO vs. website redesign: what do I need?" (comparison)
Decision: "Case study: +136% conversion increase in 10 months" (proof point)

In A/B testing

Personas give you test hypotheses. If Carlos values "clear methodology," you can test adding a "Our Process" section to the landing page. If he worries about "results being just luck," you can test adding statistical significance to case studies.

How many personas do you need

The short answer: as few as possible. The long answer:

1 primary persona: your ideal customer, the one that generates the most revenue
1-2 secondary personas: important but non-priority segments
1 negative persona (optional): the customer you don't want to attract

More than 4 personas dilute focus. If you can't prioritize, you haven't done enough research.

Conclusion

User personas aren't a creative exercise. They're a data-driven decision-making tool that should guide every aspect of your website, product, and conversion strategy. If your personas are in a PDF that nobody opens, it's time to redo them with real data.

Want to find out if your website speaks your ideal customer's language? Request a free CRO audit and analyze it in less than 2 minutes.

Ready to optimize your conversion based on real user data? Discover our CRO service and stop guessing.

Adrià Vidal is the founder of Boost. +1,000 optimization actions, +47.8% average conversion increase per client, +EUR 7.8M in additional revenue generated.