Most articles about user personas hand you a pretty template and call it a day. That’s not what designers actually need when they’re staring down a website redesign with a tight deadline and stakeholders who all think they know the user best. This guide skips the fill-in-the-blank approach and walks you through how to research, structure, and apply user personas in a way that genuinely shapes redesign decisions.
Why User Personas Matter Specifically for a Redesign
A website redesign is not the same as building from scratch. You already have traffic, behavior data, support tickets, and a brand that people interact with. Personas built for a redesign should answer one core question: who is currently struggling with the existing site, and who do we want to serve better going forward?
Generic personas like “Marketing Mary, 34, likes coffee” won’t help you decide whether the homepage hero should highlight pricing or social proof. The personas you build need to be tied to actual decisions on the redesign roadmap.
Step 1: Audit What You Already Know Before Doing Any Research
Before you talk to a single user, mine what you already have. This step alone usually surfaces 60% of what you need.
- Analytics: Which pages get the most traffic? Where do users drop off? What devices and referrers dominate?
- Heatmaps and session recordings: Where are people clicking that isn’t a link? Where do they hesitate?
- Customer support logs: What questions repeat? These are gold for understanding friction.
- Sales call notes and CRM data: What objections come up? What language do prospects use?
- Search queries on your site: What are people trying to find that they can’t?
Document recurring patterns. If 40% of support tickets are about pricing clarity, that’s a persona signal: a user who feels uncertain about commitment.
Step 2: Talk to Real Users (Yes, You Have to Do This)
No amount of analytics replaces a 30-minute conversation with someone who uses your site. Aim for 5 to 8 interviews per suspected segment. That’s enough to spot patterns without drowning in transcripts.
Who to interview
- Current customers who use the site regularly
- Customers who churned recently
- Prospects who visited but didn’t convert (pull from sales)
- Internal teams that talk to users daily (support, sales, success)
Real questions to ask (not the boring ones)
Skip “what’s your job title?” You can get that from LinkedIn. Ask things that reveal behavior and motivation:
- Walk me through the last time you visited a site like ours. What were you trying to do?
- What were you doing right before you opened the site?
- If our site disappeared tomorrow, what would you use instead and why?
- What part of the site do you avoid? Why?
- Tell me about a time the site frustrated you.
- Who else is involved in your decision when you use this site?
- What would have to be true for you to recommend us to a colleague?
Record everything (with permission). Look for verbatim phrases you can later use in microcopy and headlines.
Step 3: Cluster Patterns Into Segments
After interviews, lay out the raw insights and group them by behavior, not demographics. A 45-year-old CFO and a 28-year-old founder might behave identically when evaluating your pricing page. They belong in the same persona.
Useful clustering dimensions for a redesign:
- Goal on the site: evaluating, learning, buying, supporting, returning
- Level of awareness: first-time visitor vs. existing customer
- Decision authority: end user, influencer, or buyer
- Friction tolerance: patient researcher vs. quick decider
Aim for 3 to 5 personas maximum. More than that and your team will ignore them.
Step 4: Structure Each Persona Around Redesign-Useful Fields
Forget the stock photo and the made-up name. Here’s what actually drives design decisions:
| Field | Why it matters for the redesign |
|---|---|
| Primary job-to-be-done on the site | Drives IA and primary CTAs |
| Triggers that bring them to the site | Shapes landing page messaging |
| Key objections and doubts | Tells you what proof to surface (testimonials, specs, FAQs) |
| Information they need to decide | Determines page content and depth |
| Device and context | Influences responsive priorities and load budgets |
| Verbatim quotes | Feeds microcopy and headlines |
| Current site failures for this persona | Becomes your redesign success criteria |
Notice what’s missing: age range, favorite hobbies, marital status. Unless those facts directly change a design decision, leave them out.
Step 5: Translate Personas Into Actual Design Decisions
This is the step most teams skip, and it’s why personas end up forgotten in a Figma file. For each persona, write a short design implications section.
Example, kept deliberately concrete:
Persona: The Cautious Evaluator
Triggers: Manager asked them to find a vendor
Implications for the redesign:
- Pricing must be visible without sales contact (top objection)
- Comparison table on solution pages (they evaluate 3+ tools)
- Security and compliance info linked from the main nav, not buried in the footer
- Homepage hero should lead with a clear category statement, not a clever tagline
Do this for every persona. Now your personas are decision tools, not decoration.
Step 6: Pressure-Test Personas Against the Redesign Backlog
Take your list of planned redesign features and pages. For each one, ask: which persona benefits, and how? If a feature doesn’t serve any persona, question whether it belongs in the scope. If a persona has no features serving them, your scope has a gap.
This exercise alone has saved redesign projects from weeks of debate about whether to include a particular section.
Step 7: Keep Personas Alive After Launch
Personas decay. User behavior shifts, new segments emerge, your product changes. Build a habit:
- Revisit personas every 6 to 9 months
- Tag user feedback and support tickets by persona
- Include persona references in design critiques and PR descriptions
- Update verbatim quotes annually with fresh interviews
Common Mistakes to Avoid
- Building personas after the wireframes are done. They become rationalization, not direction.
- Using only internal opinions. Sales and support insights are valuable, but they are not user research.
- Too many personas. If you have eight, you effectively have none.
- Pretty over useful. A persona that looks great but doesn’t change a single design decision is wasted effort.
- Skipping verbatim quotes. Real language is what makes a persona feel human in a stakeholder meeting.
FAQ
How many user personas should a website redesign have?
Between 3 and 5. Fewer than 3 usually means you’re glossing over real differences in user behavior. More than 5 means your team will struggle to keep them top of mind during design work.
Can I create user personas without interviewing users?
You can, but they will be assumption-based and prone to bias. At minimum, combine analytics, support data, and conversations with customer-facing teams. Even three or four short user interviews dramatically improve persona quality.
What’s the difference between a user persona and a buyer persona?
Buyer personas focus on the purchase decision and are typically used by marketing and sales. User personas focus on behavior, goals, and friction when using the product or site. For a website redesign, you usually need user personas, though they may overlap with buyer personas if your site is sales-oriented.
Should personas include demographic data?
Only if demographics actually influence design choices. For an e-commerce site selling to older adults, age may shape font sizes and contrast. For a B2B SaaS site, age is mostly irrelevant. Include what changes decisions, leave out the rest.
How long does it take to create user personas for a website?
For a redesign, plan on 2 to 4 weeks: one week for internal data and recruiting, one to two weeks for interviews and synthesis, and a few days to structure and socialize the personas with the team.
Do AI persona generators work?
They’re useful as a starting point or for brainstorming, but they generate assumptions, not insights. Treat their output as a hypothesis to validate with real user data, never as a finished persona.
Wrapping Up
Personas are only useful when they change what you build. For a website redesign, that means anchoring them in real research, structuring them around decision-relevant fields, and explicitly translating each one into design implications. Skip the stock photos, skip the made-up biographies, and focus on what helps your team make sharper choices. That’s how personas earn their place in the redesign process.