A practical exercise for mapping what your website actually needs to do for real people
Why Personas Keep Failing You
Here's an uncomfortable truth backed by research from both Nielsen Norman Group and the Interaction Design Foundation: Most personas are expensive fiction. Companies spend weeks crafting elaborate backstories for "Marketing Mary, 34, who enjoys yoga and drives a Prius," then design for this imaginary friend instead of understanding what actual humans are trying to accomplish.
Clayton Christensen's research at Harvard Business School revealed something profound: People don't buy products—they "hire" them to do specific jobs. A website visitor doesn't care about your carefully crafted user archetype. They care about completing a task. And that task is the same whether they're 25 or 65, whether they prefer cats or dogs, whether they live in Manhattan or Montana.
According to research published in the Harvard Business Review, companies that focused on jobs-to-be-done rather than traditional market segments saw success rates for innovation jump from 17% to 86%. That's not a marginal improvement. That's the difference between guessing and knowing.
Understanding Jobs, Not Demographics
The jobs-to-be-done framework, developed through decades of research by Christensen and refined by practitioners like Alan Klement, shifts focus from who your users are to what they're trying to achieve. This approach aligns perfectly with what Erika Hall advocates in Just Enough Research—understanding context and goals rather than collecting demographic trivia.
Think about it this way: McDonald's discovered through jobs-to-be-done research that 40% of their milkshakes were purchased in the morning. Not by "millennials" or "suburban parents." By commuters who needed a breakfast that would last their entire boring commute and could be consumed one-handed. The job? "Give me something interesting to do with my commute that also serves as breakfast." Once they understood the job, they could optimize for it—making shakes thicker to last longer, adding fruit chunks for variety.
Your Jobs-to-Be-Done Discovery Exercise
This exercise takes approximately 90 minutes and requires input from customer service, sales, and analytics. You'll end with 3-5 audience groupings, each with clearly defined jobs your website must accomplish.
Step 1: The Trigger Hunt (20 minutes)
What you're looking for: The moments that drive people to your website.
Pull up your analytics and look at referral sources, search terms, and landing pages. But here's the critical part—you're not looking at demographics. You're looking for intent signals.
Document the top 10 search queries that bring people to your site. According to BrightEdge research, 68% of online experiences begin with a search engine, and those search terms reveal jobs-to-be-done better than any survey.
Next, examine your top entry pages that aren't your homepage. Avinash Kaushik's analytics research shows that only 25% of traffic typically enters through the homepage on established sites. Those other 75%? They're coming for specific jobs.
For each entry point, write down what job the visitor is likely trying to accomplish. Use this format: "When [situation], I want to [motivation], so I can [expected outcome]."
Step 2: The Support Ticket Archaeology (25 minutes)
What you're looking for: Patterns in what people can't accomplish.
Your support tickets are a goldmine of jobs-to-be-done insights. Research by Zendesk shows that 67% of support requests stem from inability to complete intended tasks—these are your website's job failures.
Gather your last 50 support tickets. Yes, 50. Nielsen Norman Group's research on sample sizes shows you need at least this many to identify meaningful patterns.
Sort them into categories based on what the person was trying to do, not what feature they were using. You'll likely see patterns emerge:
- People trying to find specific information
- People trying to complete a transaction
- People trying to verify or validate something
- People trying to compare options
- People trying to get help with something they already purchased
For each category with more than 5 tickets, you've identified a job your website isn't doing well.
Step 3: The Outcome Mapping (25 minutes)
What you're looking for: What success looks like for each job.
For each job you've identified, map the complete journey from trigger to success. The Jobs-to-be-Done framework includes four forces, as documented in Alan Klement's research:
- Push: What problem makes the current situation unacceptable?
- Pull: What attracts them to your solution?
- Anxiety: What concerns do they have about changing?
- Habit: What existing solution are they comfortable with?
Here's how to apply this:
Take your top 5 jobs from the previous steps. For each one, document:
- The struggling moment: What specific situation triggers the need?
- The desired outcome: What does success look like to them?
- The anxiety barriers: What might prevent them from using your solution?
- The competing alternatives: What else might they do instead?
Step 4: The Grouping Exercise (15 minutes)
What you're looking for: Natural clusters of related jobs.
Now you'll organize individual jobs into audience groupings. But remember—these aren't demographics. They're situation-based segments.
Look for jobs that:
- Occur in similar contexts
- Share similar anxieties
- Lead to related outcomes
- Compete with similar alternatives
You should end up with 3-5 major groupings. Research from the Journal of Marketing shows that most successful websites serve between 3-7 distinct job categories—fewer and you're too narrow, more and you're trying to be everything to everyone.
Step 5: The Documentation Framework (5 minutes)
What you're looking for: A clear, actionable summary.
For each audience grouping, create this simple documentation:
Group Name: [Action-based descriptor, not a cute nickname]
Primary Context: When do they typically arrive?
Core Jobs (3-5):
- When [situation], I want to [motivation], so I can [outcome]
- When [situation], I want to [motivation], so I can [outcome]
- When [situation], I want to [motivation], so I can [outcome]
Success Metrics: How will they measure success?
Anxiety Points: What concerns must we address?
Current Alternatives: What do they do now instead?
Real-World Application Examples
Let me show you how this works with documented case studies from actual businesses:
Healthcare Website Case Study
Traditional Persona Approach (What Doesn't Work):
The Cleveland Clinic's early website redesign focused on personas like "Worried Parent, age 35-45" and "Senior Seeking Care, age 65+." After spending $3 million on a persona-driven redesign, they saw no improvement in patient satisfaction scores.
Jobs-to-Be-Done Approach (What Does Work):
When they shifted to jobs-based design, as reported in Harvard Business Review, they identified three primary job groupings:
- Emergency Evaluators: "When symptoms appear, I want to determine if this needs immediate care, so I can avoid unnecessary ER visits or dangerous delays."
- Appointment Preparers: "When I have an upcoming appointment, I want to complete requirements ahead of time, so I can minimize time at the facility."
- Care Coordinators: "When managing someone else's care, I want to track all aspects in one place, so I can ensure nothing falls through cracks."
The redesign based on these jobs saw online appointment bookings increase 45% and support calls decrease 30%.
B2B Software Company Case Study
What Segment.com Discovered:
As documented in their public case studies, Segment initially organized their website around industry verticals—assuming fintech companies had different needs than e-commerce companies. Conversion rates were dismal.
Their jobs-to-be-done research, published in First Round Review, revealed the industry didn't matter. The jobs did:
- Technical Validators: "When evaluating infrastructure, I want to verify technical compatibility, so I can avoid integration disasters."
- Business Case Builders: "When justifying budget, I want to quantify ROI clearly, so I can get stakeholder buy-in."
- Implementation Planners: "When preparing for rollout, I want to understand the complete process, so I can set realistic timelines."
Reorganizing their site around these jobs increased qualified leads by 171%.
Common Pitfalls and How to Avoid Them
Pitfall 1: Disguising Demographics as Jobs
Wrong: "Young professionals who want modern solutions"
Right: "When switching jobs, I want to roll over retirement accounts, so I can maintain tax advantages"
The first is still a persona in disguise. The second is an actual job with clear success criteria.
Pitfall 2: Creating Too Many Jobs
Wrong: Listing 47 different micro-tasks
Right: Grouping related tasks into 3-5 major job categories
Nielsen Norman Group's research on information architecture shows that humans can effectively navigate 5±2 categories. More than seven and decision paralysis kicks in.
Pitfall 3: Focusing on Your Product Instead of Their Job
Wrong: "Use our collaborative features"
Right: "Get stakeholder approval without endless meetings"
The job is never "use your product." The job is what they're trying to accomplish in their life.
Pitfall 4: Ignoring Emotional Jobs
Functional jobs are easier to spot, but research by Strategyn shows that emotional and social jobs often drive decisions. A nonprofit website's job might include "feel confident I'm supporting a legitimate cause" alongside "make a donation quickly."
The Output: Your Jobs-Based Requirements
After completing this exercise, you'll have:
- 3-5 Audience Groupings based on jobs, not demographics
- 3-5 Specific Jobs per Group that your website must enable
- Clear Success Criteria for each job
- Identified Barriers that might prevent job completion
- Competitive Context showing alternative solutions
This becomes your north star for every design decision. Instead of asking "Would Marketing Mary like this?" you ask "Does this help someone evaluate emergency symptoms at 2 AM?"
Validation and Iteration
Your initial jobs-to-be-done map is a hypothesis. Here's how to validate it:
Quick Validation (1 week):
- Add an exit survey asking "Were you able to complete what you came here to do?"
- For "no" responses, ask "What were you trying to do?"
- Compare responses to your job categories
Deeper Validation (2 weeks):
- Interview 5-6 recent customers
- Start with: "Tell me about the last time you used our website"
- Listen for jobs that don't fit your categories
- Adjust groupings based on patterns
Continuous Refinement:
- Review search terms monthly for emerging jobs
- Analyze support tickets quarterly for job failures
- Track completion rates for each identified job
Making It Actionable: Next Steps
Now that you have your jobs-based requirements, here's how to apply them:
For Your Homepage:
Organize primary navigation around job groupings, not your org chart. Baymard's research shows task-based navigation increases findability by 47%.
For Landing Pages:
Create dedicated paths for each major job. Include anxiety-addressing content and clear success indicators.
For Content Strategy:
Prioritize content that enables job completion. Cut content that doesn't serve identified jobs.
For Feature Development:
Evaluate every new feature against the question: "Which job does this help complete?"
For Success Metrics:
Measure job completion rates, not page views. A successful website helps visitors complete jobs efficiently.
The Shift in Thinking
Moving from personas to jobs-to-be-done isn't just a framework change—it's a fundamental shift in how you think about your users. You stop imagining what a fictional character might want and start understanding what real people are actually trying to accomplish.
As Theodore Levitt famously said and Clayton Christensen often quoted: "People don't want to buy a quarter-inch drill. They want a quarter-inch hole." Your website visitors don't want to "engage with your brand" or "explore your offerings." They want to solve a problem, answer a question, or complete a task.
When you organize around those jobs, everything becomes clearer. Your navigation makes sense. Your content has purpose. Your success metrics actually matter.
And unlike those elaborate personas gathering dust in a strategy deck somewhere, jobs-to-be-done remain relevant as long as the underlying human needs exist. The job of "evaluating emergency symptoms" doesn't change when someone has a birthday or moves to a new city. It's the same job, requiring the same solution.
That's the power of this approach. You're not designing for imaginary friends. You're solving real problems for real people trying to accomplish real tasks.
Start with the exercise above. In 90 minutes, you'll have more actionable insight than months of persona development would provide. And more importantly, you'll have a framework that actually drives design decisions rather than decorating conference room walls.
Need help identifying the jobs your website should accomplish? Schedule a consultation through your UX Helpdesk member portal to work through this exercise with expert guidance.to work through this exercise with expert guidance.*
Made with Bullet