JTBD vs. Personas: Why Traditional Segmentation Fails

The Six Personas That Were All Wrong

A medical device company — a mid-size European firm with strong positions in surgical instruments — came to us with a problem. They had invested six months and a significant budget developing buyer personas for their flagship product line. Six detailed profiles: “Efficient Eva,” “Research-Driven Rainer,” “Budget-Conscious Barbara,” and three others. Each persona had a name, a photo, a backstory, demographics, psychographics, goals, frustrations, and preferred information channels.

The personas were beautifully designed. The product team had them printed on large format displays mounted on their office walls. Every sprint review referenced them. Feature prioritization meetings began with “which persona does this serve?”

There was only one problem: the personas were fiction. And the product decisions made based on them were consistently wrong.

When we ran a JTBD/ODI study on the same product line, surveying 340 surgeons and OR staff across four European markets, we found something the personas could not have revealed: the customer base segmented into three distinct groups — not based on demographics, title, hospital type, or years of experience, but based on patterns of unmet needs. Two surgeons who matched “Efficient Eva” perfectly — same age, same specialty, same hospital size, same stated goal of efficiency — fell into different segments with radically different priorities. One was underserved on speed-related outcomes. The other was underserved on precision-related outcomes. The persona grouped them together. The JTBD analysis showed they needed different product configurations.

This is not an isolated case. It is the norm. And it reveals a fundamental flaw in how most product teams segment their markets.

What Personas Get Right

Before we dismantle personas, let us acknowledge what they do well.

Personas humanize the customer. In organizations where engineering teams are three degrees removed from the end user, a well-constructed persona reminds people that they are building for actual humans with real constraints and motivations. This is genuinely valuable — particularly in large organizations where product decisions can become abstract exercises in feature comparison.

Personas also serve as a shared reference point. When a product manager says “this feature is for Efficient Eva,” everyone in the room has a mental model of who that customer is. It creates alignment, even if that alignment is based on a simplification.

And personas can encode useful contextual information: typical workflows, technology ecosystems, organizational constraints, buying authority. These contextual details inform design decisions and go-to-market strategy.

The problem is not that personas are useless. The problem is that they segment the market on the wrong axis — and teams make high-stakes product investment decisions based on that segmentation.

The Fundamental Flaw: Segmenting by Who, Not by Why

Personas segment by demographics, psychographics, and behaviors — who the customer is. JTBD segments by what the customer is trying to accomplish and where current solutions fall short — why they buy.

This distinction sounds academic until you see its consequences in real product decisions.

The Demographic Trap

Consider two chief surgeons at comparable university hospitals in Germany. Both are 48 years old, lead departments of 15-20 staff, perform 200+ procedures per year, and describe themselves as “early adopters” of surgical technology. By any persona framework, they are the same customer.

But Surgeon A’s primary frustration is setup time. She works in a hospital with tight OR scheduling, and she measures every product by how quickly her team can prepare for the next case. Her underserved outcomes cluster around speed: minimize the time to configure the instrument, minimize the time to verify calibration, minimize the time to transition between procedure types.

Surgeon B’s primary frustration is tactile feedback. He performs complex reconstructive procedures where millimeter precision determines patient outcomes. His underserved outcomes cluster around control: minimize the likelihood of unintended tissue damage, minimize the variability of instrument response under different tissue conditions, minimize the force required to make fine adjustments.

A persona sees one customer. JTBD sees two segments with different priorities, requiring different product configurations, different messaging, and potentially different pricing models. The persona-based roadmap will partially satisfy both surgeons and fully satisfy neither. The JTBD-based roadmap can target each segment with precision.

Why Persona-Based Segmentation Persists

If personas are flawed as a segmentation tool, why do they remain the default in most organizations? Three reasons:

1. They Are Easy to Create

A skilled UX researcher can build persona profiles in 2-4 weeks with 12-15 interviews. The output is tangible, visually appealing, and immediately shareable. JTBD/ODI research takes 8-12 weeks and requires quantitative survey methodology. In organizations optimizing for speed of insight delivery, personas win on convenience.

2. They Feel Intuitive

Humans think in terms of people, not in terms of need patterns. “This product is for busy surgical department heads” is easier to internalize than “this product targets the segment where 67% of respondents rate speed-related outcomes as highly important and less than 25% satisfied.” The first statement tells a story. The second is a data point. Stories are more memorable. Data is more accurate.

3. They Avoid Hard Conversations

Persona-based segmentation rarely forces trade-offs. If you have six personas, you can claim that your product serves all of them. JTBD-based segmentation forces you to choose: which segment are we targeting first? Which underserved outcomes are we prioritizing? These are difficult strategic conversations that many product teams prefer to defer.

Traditional segmentation methods fail because they group customers by attributes that do not predict purchasing behavior. Age, income, title, and company size do not determine what customers need from a product. Patterns of unmet needs do. Until you segment on unmet needs, you are guessing.

The JTBD Alternative: Needs-Based Segmentation

JTBD segmentation works differently. Instead of grouping customers by observable attributes, it groups them by the outcomes they find most important and least satisfied.

Here is how it works in practice:

Step 1: Capture All Desired Outcomes

Through qualitative interviews (15-30 participants), identify the 50–150 outcomes customers use to evaluate how well the job is getting done. These outcomes span all stages of the job map and all three dimensions (functional, emotional, social). For a detailed guide to conducting these interviews effectively, see our JTBD interview guide.

Step 2: Quantify Importance and Satisfaction

Survey a representative sample (200-600 respondents) on each outcome. For each outcome, ask: “How important is this to you?” and “How satisfied are you with current solutions on this dimension?” Use a consistent scale (typically 1-10).

Step 3: Calculate Opportunity Scores

For each outcome, calculate the opportunity score using Ulwick’s formula: Opportunity = Importance + max(Importance - Satisfaction, 0). Outcomes with high importance and low satisfaction have the highest opportunity scores — these are underserved needs.

Step 4: Cluster by Need Patterns

Using statistical clustering methods (k-means, hierarchical clustering) on the outcome-level data, identify groups of customers who share similar patterns of unmet needs. These are your segments — defined not by who the customers are, but by what they need.

The Result: Segments That Predict Behavior

The segments that emerge from this process are remarkably predictive. Customers in the same JTBD segment respond to the same product features, the same messaging, and the same value propositions — even when their demographics are completely different. This is because the segmentation is based on the actual drivers of purchasing behavior (unmet needs), not on proxies for those drivers (demographics).

In our work with the medical device company mentioned at the start of this article, the three JTBD-based segments correlated with market share gain in the subsequent product launch at more than double the rate of the persona-based predictions. The segment that the personas called “low priority” (because it did not match any persona profile closely) turned out to be the highest-growth segment.

Where Personas Still Have a Role

I am not arguing that you should burn your persona documents. Personas serve legitimate functions — just not the one most teams use them for.

Communication and Empathy

Personas are excellent tools for building empathy with the customer within engineering and design teams. They make abstract market data concrete. After you have identified your JTBD-based segments, you can create persona-like profiles for each segment — but now the profile is anchored to real need patterns rather than demographic assumptions.

Content and Channel Strategy

Personas encode useful information about preferred communication channels, content consumption habits, and decision-making processes. This information is valuable for marketing and sales enablement, even if it should not drive product strategy.

Stakeholder Communication

When presenting to senior leadership or boards, persona-style profiles can make segment descriptions more accessible. Not every executive needs to understand cluster analysis. They do need to understand who the customer is and what they need.

The key principle: use JTBD for strategic decisions (what to build, where to invest, how to segment). Use personas for tactical execution (how to communicate, where to reach customers).

A Head-to-Head Comparison

The B2B Dimension: Where Personas Fail Most Visibly

In B2B markets, the limitations of personas are amplified. A B2B purchasing decision involves multiple stakeholders — the user, the buyer, the maintainer, the compliance officer, the executive sponsor — each with different jobs and different unmet needs.

Personas in B2B typically represent these roles: “Operator Oliver,” “Procurement Petra,” “C-Suite Carsten.” But the real complexity is not that different roles have different needs (that is obvious). The real complexity is that people in the same role, at similar companies, in the same industry, have different patterns of unmet needs. Two procurement managers at comparable manufacturers may care about completely different outcomes. One is underserved on total cost of ownership transparency. The other is underserved on supplier responsiveness during technical evaluation.

JTBD handles this complexity naturally. By defining the job for each stakeholder role and quantifying outcomes separately, you get a multi-dimensional map of the buying committee’s needs. This is why JTBD for B2B is arguably more powerful than JTBD for consumer markets — the complexity that makes B2B hard is exactly the complexity that JTBD is designed to handle.

Making the Transition: Practical Steps

If you are currently using personas and want to move toward JTBD-based segmentation, here is a practical transition path:

Step 1: Do Not Discard Personas Immediately

Keep your existing personas for marketing and communications purposes. Do not create organizational disruption by announcing that personas are “wrong.” Instead, introduce JTBD as an additional input to product strategy.

Step 2: Run a JTBD Pilot Alongside Existing Personas

For one product line, run a full JTBD/ODI study. Compare the segments that emerge from the data to your existing persona definitions. The differences will be instructive — and will build the case for JTBD-based segmentation without requiring a top-down mandate.

Step 3: Map JTBD Segments to Reachable Attributes

One legitimate concern about JTBD segmentation is reachability. You cannot buy a media list of “customers who are underserved on precision outcomes.” After identifying your JTBD segments, run a discriminant analysis to find observable attributes (industry, company size, application type, technology generation in use) that correlate with segment membership. This creates a bridge between JTBD insights and practical targeting.

Step 4: Evolve Personas into Segment Profiles

Over time, replace your demographic personas with segment profiles that are anchored in unmet needs but enriched with contextual information. “Segment A: Precision-Driven Surgeons” is more actionable than “Efficient Eva” because it is tied to quantified outcomes and verified through data. These segment profiles then become the foundation for designing a value proposition that speaks directly to each segment’s most critical underserved outcomes.

The best transitions happen when the product team experiences the contrast firsthand. Run one roadmap review with persona-based inputs and one with JTBD-based inputs. The difference in specificity and confidence is immediately apparent. I have never seen a product team go back to personas for strategic decisions after experiencing JTBD-based segmentation.

Frequently Asked Questions