Back to Insights
Customer Research7 min readApril 14, 2026

Why 90% of Customer Personas Are Useless (And What to Do Instead)

Most persona documents gather dust because they're built on assumptions, not evidence. Here's the systematic approach that actually changes how you build and market.

The problem with most personas

You've seen them. A template with a stock photo, a name like "Marketing Mary," an age range, and a list of interests including "coffee" and "work-life balance." These documents are created in a sprint, added to a Notion page, and never opened again.

They fail because they're built backwards. Teams start with demographics (age, income, location) and work toward behavior. Real buying behavior is driven by the opposite: the job someone is trying to accomplish, the friction blocking them, and the outcome they're convinced is possible.

The three root causes of useless personas

1. They're built from assumptions, not evidence.

The most common persona creation process is a team workshop where people brainstorm who they think their buyer is. This produces a document that reflects the team's internal beliefs, not market reality. The personas look professional but confirm existing biases.

2. They describe who buys — not why.

Knowing that your buyer is a 35-year-old female marketing manager earning $90k is nearly useless. The actionable question is: what is she trying to accomplish that she hasn't been able to do yet? This is the Jobs-to-be-Done (JTBD) framework, and it's the difference between a persona that informs strategy and one that decorates a slide deck.

3. They don't segment strategically.

Most companies have one persona. Real market strategy requires at least three: a Cash Cow (high LTV, easy to close), a Growth Segment (scalable but requires more nurturing), and a Negative Persona (who looks like your buyer but isn't — and costs you in support, churn, and ad waste).

What evidence-based personas actually look like

A good persona document includes:

The practical fix

Stop building personas in workshops. Build them from evidence:

1. Customer interviews — 5-8 interviews with your best customers. Ask: "What were you trying to do before you found us? What made you look for a solution? What almost stopped you from buying?"

2. Win/loss analysis — go through your last 20 closed deals and 10 churned accounts. What did winners have in common that losers didn't?

3. AI-grounded synthesis — tools like Zamicus run your product description through a multi-agent pipeline that extracts industry-specific buyer profiles grounded in market research, not team brainstorming.

The output from even 5 good customer interviews, processed through a structured JTBD framework, will outperform any workshop-generated persona document.

What to do with a good persona

A persona is only useful if it changes a decision. The test is simple: look at your persona document and ask — does this change what I write in the headline of my landing page? Does it change which channel I invest in first? Does it change what I offer in month 1 vs month 3?

If the answer is no, the persona is decorative. Start over.

No More Guessing

See your customers clearly
for the first time.

Generate your first persona in 3 minutes. No data required. No credit card. Just clear answers.