Jobs to Be Done (JTBD): How to Build Products People Actually Need

Jobs to Be Done (JTBD): How Product Managers Build What Customers Actually Need
Product Management • Product Discovery • Customer Understanding

One of the biggest mistakes in product management is asking:

“What feature should we build?”

before asking:

“What is the customer actually trying to accomplish?”

Jobs to Be Done, commonly called JTBD, is a way of understanding customers based on the progress they are trying to make rather than simply looking at who they are.

It helps Product Managers understand the real reason behind a purchase, behaviour or product decision.

What Is Jobs to Be Done?

In simple terms, JTBD says:

People don’t simply buy products. They “hire” products to help them get a job done.

The “job” is the outcome the person is trying to achieve.

The product is simply one of the ways they can achieve that outcome.

For example, someone does not necessarily buy a drill because they love drills.

They buy a drill because they want a hole.

The Drill Analogy

Imagine I ask 100 people:

“Why did you buy a drill?”

Many might answer:

  • “It has a powerful motor.”
  • “The battery lasts a long time.”
  • “It is lightweight.”
  • “It is from a good brand.”

Those are product attributes.

But the underlying job might be:

“I need to securely put something on my wall.”

Suddenly, the product problem looks different.

Perhaps the customer does not really care about owning a drill. They care about hanging a shelf quickly and safely.

The real competition may therefore not be another drill.

It could be:

  • Adhesive hooks.
  • Command strips.
  • A professional handyman.
  • A different mounting system.

This is the power of JTBD.

JTBD Changes How a Product Manager Thinks

A traditional product conversation might sound like:

“Customers are asking for a dashboard. Let’s build a dashboard.”

A JTBD-oriented PM would ask:

“What are customers trying to accomplish that makes them ask for a dashboard?”

Maybe the customer actually wants:

  • To understand what is happening in the business.
  • To identify problems quickly.
  • To report performance to management.
  • To make a decision without manually combining five reports.

Now the product team has much more flexibility.

The solution might be a dashboard.

But it might also be:

  • An automated report.
  • An alert system.
  • An AI assistant.
  • A weekly email summary.
  • A mobile notification.
JTBD prevents the team from confusing the customer’s requested solution with the customer’s actual problem.

Job vs Need vs Pain Point vs Feature

These concepts are often mixed together.

Concept Meaning Example
Job What the customer is trying to accomplish. “I need to follow up with patients on time.”
Need Something required to accomplish the job. “I need to know which patients need follow-up.”
Pain Point What makes the job difficult today. “Staff manually check spreadsheets and often miss tasks.”
Feature The product mechanism used to solve the problem. “Automated follow-up task creation.”

The mistake is starting at the bottom.

Strong product discovery usually works upwards:

Customer Job → Pain → Desired Outcome → Solution → Feature

Functional, Emotional and Social Jobs

Jobs are not always purely functional.

1. Functional Job

What does the customer practically need to accomplish?

Example:

“I need to schedule appointments quickly.”

2. Emotional Job

How does the customer want to feel?

Example:

“I want to feel confident that I haven’t forgotten an important patient.”

3. Social Job

How does the customer want to be perceived by others?

Example:

“I want my clinic to look organised and professional.”

A strong product can address all three.

A Simple Real-Life Example: Why Do People Buy Coffee?

Imagine two people buying coffee at 8:00 AM.

They order the same coffee.

But they may have completely different jobs.

Person A

Their job is: “I need something quick before I get to work.”

They care about:

  • Speed.
  • Convenience.
  • Predictability.

Person B

Their job is: “I want somewhere comfortable to sit and start my morning.”

They care about:

  • Atmosphere.
  • Comfort.
  • Wi-Fi.
  • Environment.

Same product category.

Different jobs.

Therefore, different product experiences.

Real Product Example: Netflix

Consider the job behind streaming entertainment.

A simplistic view might be:

“People want movies and TV shows.”

But the underlying jobs can be much broader:

  • “I want to relax after work.”
  • “I want something entertaining without spending time deciding.”
  • “I want something my family can watch together.”
  • “I want to keep up with a show everyone is talking about.”

This changes how we think about the product.

Recommendation systems, search, personalised rows and continue-watching features are not valuable simply because they are technically impressive.

They help users make progress on the underlying job:

“Help me quickly find something I actually want to watch.”

The job is therefore bigger than the individual feature.

Real Product Example: Uber

A customer does not necessarily want an Uber.

The customer wants to get from one place to another.

Their job could be:

“Help me get to my destination safely, predictably and without the hassle of finding transport myself.”

This job opens up many product considerations:

  • Finding a driver.
  • Estimating arrival time.
  • Showing the route.
  • Providing price information.
  • Handling payment.
  • Providing safety features.
  • Helping users recover when something goes wrong.

Again, the job is not “book an Uber.”

Booking the Uber is simply one step towards achieving the desired outcome.

JTBD and Product Discovery

JTBD becomes particularly powerful during product discovery.

Before building anything, I want to understand:

  1. Who has the problem?
  2. What are they trying to accomplish?
  3. When does the problem happen?
  4. What triggers them to solve it?
  5. How do they solve it today?
  6. What makes the current solution frustrating?
  7. What outcome would they consider successful?

How I Would Conduct a JTBD Interview

If I were the Product Manager, I would not start by asking:

“Would you use an AI feature that automatically does this?”

That question introduces my solution into the conversation.

Instead, I would investigate the customer’s actual behaviour.

Step 1: Understand the situation

“Tell me about the last time you had to do this.”

Step 2: Understand the trigger

“What happened that made you decide to do it?”

Step 3: Understand the current process

“Can you walk me through exactly what you did?”

Step 4: Identify friction

“What was the most frustrating part?”

Step 5: Understand alternatives

“What did you use before this?”

Step 6: Understand the desired outcome

“If this worked perfectly, what would be different?”

Ask About Real Behaviour, Not Hypothetical Opinions

This is extremely important.

There is a big difference between:

“Would you use this feature?”

and:

“Tell me about the last time you had this problem.”

People are often poor at predicting what they will do in the future.

Their existing behaviour provides stronger evidence.

Healthcare Case Study: Automating Patient Follow-Ups

Let’s imagine we are developing a healthcare automation product.

The initial request from clinic managers is:

“We need an automated patient follow-up feature.”

A feature-focused team might immediately start designing:

  • Automation rules.
  • Templates.
  • Scheduling settings.
  • Notifications.
  • Patient segmentation.

But as the PM, I would first investigate the job.

Step 1: Interview Clinic Staff

I might speak with:

  • Practice managers.
  • Receptionists.
  • Administrators.
  • Clinicians.
  • Customer support teams.
  • Operations managers.

Step 2: Understand the Current Workflow

I discover that staff currently:

  1. Review patient records.
  2. Identify patients requiring follow-up.
  3. Check spreadsheets.
  4. Manually create tasks.
  5. Send messages.
  6. Record the outcome.
  7. Repeat the process every day.

Step 3: Identify the Real Job

The job is not simply:

“Send automated messages.”

The broader job might be:

“Make sure the right patients receive the right follow-up at the right time without staff having to manually track every case.”

Step 4: Identify the Pain

Staff explain that:

  • They spend significant time checking lists.
  • Some tasks are forgotten during busy periods.
  • Different staff members manage follow-ups differently.
  • Managers cannot easily see outstanding work.

Step 5: Define the Desired Outcome

The desired outcome could be:

“Reduce the amount of manual tracking required while increasing the reliability of patient follow-up.”

Step 6: Design the Solution

Only now would I start thinking about the product.

Potential functionality could include:

  • Automatic identification of follow-up requirements.
  • Configurable workflow rules.
  • Staff approval before sensitive communications are sent.
  • Patient communication templates.
  • Task creation.
  • Escalation when a follow-up is not completed.
  • Audit history.
  • Reporting for managers.

Notice what happened.

We did not start with:

“Let’s build an AI automation feature.”

We started with:

“What job are healthcare workers trying to get done?”

JTBD Helps Define the MVP

One of the biggest benefits of JTBD is that it can stop teams from overbuilding.

Suppose the healthcare team proposes:

  • AI-generated messages.
  • Advanced analytics.
  • Custom workflow builder.
  • Predictive patient scoring.
  • Mobile application.
  • Voice assistant.
  • Advanced dashboards.

But the core job is simply:

“Make sure important follow-ups are not forgotten.”

The MVP might therefore only need:

  1. Identify a follow-up task.
  2. Create the task automatically.
  3. Notify the responsible staff member.
  4. Track whether it was completed.

Everything else can be tested later.

From JTBD to Product Requirements

JTBD does not replace requirements.

It gives you a better foundation for writing them.

JTBD Finding Product Requirement
Staff need to know which patients require follow-up. System identifies eligible follow-up cases.
Staff need confidence before communication is sent. Staff can review and approve automated communications.
Managers need visibility into outstanding work. Dashboard shows pending and completed follow-ups.
Staff need to avoid repetitive manual tracking. System automatically creates and updates follow-up tasks.

JTBD and User Stories Are Not the Same

User stories are often written as:

“As a clinic administrator, I want to automatically create follow-up tasks so that I don’t have to track them manually.”

This is useful.

But JTBD goes deeper.

JTBD asks why the user needs to accomplish this in the first place.

The underlying job could be:

“Ensure important patient follow-ups happen reliably without adding more administrative workload.”

That deeper understanding can influence the entire product.

JTBD Can Also Help With Prioritisation

Imagine customers request ten different features.

Instead of voting on the features individually, I would map them back to the underlying jobs.

Feature Request Underlying Job
Better dashboard Understand what requires attention.
More notifications Ensure important work is not forgotten.
Bulk actions Complete repetitive work faster.
AI summaries Understand information without reading everything manually.

Now you can see that several feature requests may actually represent the same underlying job.

That creates opportunities to solve the problem more effectively than simply building every requested feature.

JTBD Helps You Understand Competition Differently

One of the most powerful parts of JTBD is that your competition is not necessarily another company selling the same type of product.

Your customer may be “hiring” any solution that gets the job done.

That could be:

  • A competitor’s product.
  • An Excel spreadsheet.
  • Email.
  • A notebook.
  • A WhatsApp group.
  • A human employee.
  • Doing nothing.
Example:

If your healthcare automation product competes with a receptionist manually maintaining a spreadsheet, the spreadsheet is part of your competitive landscape.

The “Switching” Moment

JTBD research becomes particularly interesting when you understand why someone changes from one solution to another.

Ask:

  • What were you using before?
  • Why did you start looking for something else?
  • What happened that triggered the search?
  • What alternatives did you consider?
  • Why did you choose this solution?
  • What nearly stopped you from switching?

These questions reveal what actually matters at the point of purchase.

A Practical JTBD Framework for Product Managers

1. Identify the Situation

When does the problem occur?

2. Identify the Job

What is the customer trying to accomplish?

3. Understand the Current Solution

How do they accomplish it today?

4. Identify Friction

What makes the current solution difficult, slow, expensive or unreliable?

5. Define the Desired Outcome

What would success look like?

6. Generate Solutions

What are the different ways we could solve the job?

7. Define the MVP

What is the smallest solution that can accomplish the job?

8. Validate

Are customers actually using it and achieving the intended outcome?

How JTBD Connects to the Product Lifecycle

Lifecycle Stage How JTBD Helps
Discovery Understand the problem customers are trying to solve.
Definition Define the target outcome.
Design Explore different ways to solve the job.
Development Keep engineering focused on the underlying customer outcome.
Launch Position the product around the customer’s job and value.
Post-launch Measure whether customers are successfully completing the job.
Growth Identify adjacent jobs and customer segments.
Decline Understand whether the underlying job still exists and matters.

What If Customers Ask for a Feature?

This happens constantly in product management.

A customer says:

“We need an export to Excel button.”

The worst response is automatically:

“Okay, let’s add Excel export to the backlog.”

Instead, ask:

“What are you trying to do with the data after you export it?”

The customer might say:

“I need to send a weekly report to my manager.”

Now we understand the job.

Maybe Excel export is the right solution.

But maybe an automated weekly report is much better.

The customer requested a feature. The PM discovered the job.

What JTBD Does Not Mean

JTBD does not mean:

  • Ignore what customers ask for.
  • Never build requested features.
  • Always choose the simplest solution.
  • Never use personas.
  • Never look at demographics.
  • Only conduct interviews.

It means that customer requests should be understood in context.

A feature request is a valuable signal.

It is simply not always the complete problem statement.

Common JTBD Mistakes

1. Defining the Job as a Feature

Bad:

“The job is to use our dashboard.”

Better:

“The job is to quickly understand what requires attention.”

2. Making the Job Too Broad

“Help people live better” is too broad to guide product decisions.

A useful job needs enough specificity to understand the context and desired outcome.

3. Asking Leading Questions

Avoid:

“Would an AI assistant make this easier?”

Instead:

“How do you currently do this?”

4. Only Talking to Existing Customers

Existing customers can tell you why they chose your product.

But lost customers, rejected prospects and people using alternative solutions can reveal equally important information.

5. Assuming One Customer Has One Job

Customers can have multiple jobs depending on the situation.

A clinic manager may want:

  • Operational efficiency.
  • Visibility.
  • Compliance.
  • Better patient experience.
  • Lower administrative workload.

Understanding which job is most important in a particular situation is critical.

JTBD and AI Products

JTBD is particularly useful when building AI products.

AI makes it tempting to start with the technology:

“We can build an AI agent that does X.”

But the better question is:

“What customer job would become significantly easier if this capability existed?”

For example:

Technology Potential Job
LLM Understand large amounts of information quickly.
AI summarisation Extract important information without reading everything.
AI agent Complete a multi-step workflow without manual coordination.
Recommendation model Make a decision when there are too many options.
Predictive model Identify a likely future outcome early enough to act.

The technology is not the product.

The technology becomes valuable when it helps someone make meaningful progress on a real job.

How I Would Use JTBD as a Product Manager

If I joined a new product team tomorrow, I would use JTBD as part of my discovery process rather than treating it as a one-time exercise.

  1. Understand the company’s target customers.
  2. Review existing product analytics.
  3. Review support tickets and customer feedback.
  4. Speak with customers.
  5. Observe how they currently solve the problem.
  6. Identify the important jobs.
  7. Map pain points and desired outcomes.
  8. Identify existing alternatives.
  9. Prioritise the most valuable problems.
  10. Generate multiple possible solutions.
  11. Build the smallest useful solution.
  12. Measure whether customers achieve the intended outcome.

A Complete Example: From Customer Problem to Product

Situation

A healthcare clinic manages hundreds of patient follow-ups every week.

Current Behaviour

Staff use spreadsheets, reminders and manual task lists.

Pain

The process is repetitive and some follow-ups are missed during busy periods.

Job

“Ensure important patient follow-ups happen reliably without requiring staff to manually track every case.”

Desired Outcome

Less manual tracking, fewer missed follow-ups and better visibility.

Possible Solutions

  • Automated task creation.
  • Patient follow-up dashboard.
  • Automated reminders.
  • Workflow automation.
  • AI-assisted prioritisation.

MVP

Automatically identify eligible follow-ups, create tasks and notify the responsible staff member.

Success Metrics

  • Reduction in manual tracking time.
  • Follow-up completion rate.
  • Task completion time.
  • Staff adoption.
  • Retention.
  • Customer satisfaction.

The Bigger Lesson

Product Managers are often surrounded by feature requests.

Customers ask for dashboards.

Sales asks for integrations.

Engineering suggests technical improvements.

Leadership asks for AI.

Competitors launch new features.

The Product Manager has to bring all of this back to one fundamental question:

“What problem are we actually trying to solve for the customer?”

JTBD provides a structured way to answer that question.

Final Framework

Situation → Job → Pain → Desired Outcome → Solution → Feature → Behaviour → Outcome

This sequence is important.

If you start with the feature, you are guessing.

If you start with the job, you have a reason for building.

And that changes the entire product development process.

Great Product Managers do not simply build what customers ask for. They understand what customers are trying to accomplish, why the current solutions are not good enough, and then find the best way to help them make progress.

The PM Mindset

Don’t ask: “What feature should we build?”

Ask: “What job is the customer trying to get done?”

Then ask: “What is stopping them from doing it well today?”

Then: “What is the simplest, most valuable way we can help?”

Leave a Reply

Your email address will not be published. Required fields are marked *