Jobs to Be Done

What it is

Jobs to Be Done (JTBD) is a framework for understanding why people actually choose a product. Popularized by Harvard's Clayton Christensen, its central claim is deceptively simple: customers don't buy products, they hire them to get a job done. The "job" is the progress a person is trying to make in a particular situation, and it explains behavior that demographics and feature lists cannot.

The power of the framing is that it reveals your true competition. Once you know the job, you see that you're not just competing with products that look like yours, you're competing with every other way a person could make that same progress, including the option of doing nothing at all.

This page is the practical summary. For the full story, including how JTBD explains the exact answers on your PMF survey, read the deep dive: Jobs to Be Done: why your users say "very disappointed" (or don't).

When to use it

• You're deciding what to build. JTBD keeps the roadmap anchored to progress users want, not features that demo well.
• Your feature bets keep missing. If shipping more hasn't moved retention or your score, you may be improving the wrong job.
• You're defining who your product is for. The job comes first, the ICP falls out of it.
• Users churn for reasons you can't explain. The job reveals whether they ever had one you solved, or hired you by mistake.

How to apply it

1. Interview recent switchers. Talk to people who just started or just stopped using your product, and have them walk you through what actually happened, moment by moment.
2. Find the moment of struggle. Locate the specific situation that pushed them to look for a solution. The job lives in that moment, not in a job title or an age bracket.
3. Write the job as progress. State it in the user's own words as the progress they're trying to make, for example "help me explore lots of design directions fast," not "AI design tool."
4. Map the competing forces. Note what pulls them toward you and what holds them back, including inertia and the habit of doing nothing.
5. Design for the job. Shape the product, onboarding, and messaging around making that exact progress faster and more reliable than any alternative.

Example

Christensen's milkshake study is the canonical illustration. A fast-food chain wanted to sell more milkshakes and tried the obvious moves: thicker, cheaper, more flavors. Sales didn't budge. Then the team watched who actually bought them. A large share were bought alone, early in the morning, by commuters facing a long, dull drive. They weren't buying dessert. They were hiring something easy to drink one-handed that would last the commute and hold off hunger until lunch. The real competitors weren't other milkshakes, they were bagels, bananas, and boredom. Once the job was clear, the product could finally be improved in ways that mattered.

Common mistakes

• Mistaking features for jobs. "Users want dark mode" is a feature request. The job is whatever they're trying to accomplish that dark mode would help with.
• Anchoring on demographics. Two people with identical profiles can hire your product for completely different jobs. The situation matters more than the persona.
• Only interviewing happy users. The people who switched away often reveal the job most clearly, by describing where you failed to help them make progress.
• Stopping at the functional job. Jobs have emotional and social dimensions too. "Look competent to my team" is often doing more work than the functional task.

How it connects to your PMF score

The Sean Ellis survey asks how a user would feel if they could no longer use your product. That answer is really a proxy for one question: does this product do a real job well enough that losing it leaves a gap? A strong job-fit produces "very disappointed." A missing job produces "not disappointed," regardless of how polished the product is.

That's why two users with identical usage can land in different buckets. Usage tells you what someone did. The job tells you why, and only the why predicts whether they'd fight to keep you. Track the very disappointed segment over time, and you're really tracking how well you're doing the job.