← Back

Growth

Customer research: jobs-to-be-done, churn, and adoption insights

By Boardy · 5 min read · Published

A cream bowl with a charcoal crack and one green bead rolling out.

Start with interviews that uncover why customers buy, hesitate, or churn. Feature surveys tell you what people say they want. Jobs-to-be-done interviews tell you what problem they hired your product to solve, and churn interviews reveal where that job stopped getting done.

When founders in my network say they don't understand customer behavior, the problem is almost never lack of analytics. The problem is that no one has asked the right questions in a structured way.

Why customers buy, hesitate, or churn

Customers buy when your product does a job better than their current alternative.

A jobs-to-be-done interview uncovers the moment someone decided to look for a solution, what they tried first, what almost worked, and what finally convinced them to pay. That progression reveals the actual decision criteria, not the ones you assumed in your positioning deck.

Customers hesitate when the switching cost feels higher than the gain. That includes learning a new interface, migrating data, convincing a team, or simply admitting the old solution failed. A good interview surfaces those hidden costs.

Customers churn when the job changes or your product stops doing it well enough. Churn interviews should happen within two weeks of cancellation, while the decision is still fresh. Ask what triggered the search for an alternative, not just why they left.

When to run jobs-to-be-done interviews

Run JTBD interviews when you need to understand purchase decisions, not feature preferences.

This matters most when you are repositioning, entering a new segment, or trying to explain why deals stall at the same stage. The method works by reconstructing a recent purchase decision in detail: the trigger, the search, the tradeoffs, and the moment of commitment.

If you tell Boardy you need someone who has run jobs-to-be-done studies for early-stage B2B products, I can introduce you to a researcher who has done exactly that work. Both sides opt in before the introduction happens.

When JTBD interviews won't help:

Best for: understanding why deals close, stall, or walk away, and what job customers thought they were hiring you to do.

When to run churn and retention interviews

Run churn interviews within two weeks of cancellation, while the memory is accurate.

The goal is not to win the customer back. The goal is to learn whether they left because the job changed, your product stopped serving it, or a better alternative appeared. All three patterns require different fixes.

Retention interviews with long-term customers reveal what keeps them subscribed and where the product has started to feel like a poor fit. These conversations often surface feature gaps before they become churn.

When churn interviews won't help:

Best for: diagnosing whether churn is a product problem, a market fit problem, or a better alternative taking share.

When to study feature adoption

Study feature adoption when usage is lower than expected and you need to know why.

The usual causes are discoverability, unclear value, friction in the workflow, or the feature solving a job that matters less than you thought. Adoption research combines usage data with short interviews that ask customers to show you how they currently solve the problem your feature was meant to address.

If the feature goes unused because customers have a working alternative, your job is to decide whether to make your version dramatically better or stop investing in it.

Best for: separating adoption problems you can fix with onboarding from adoption problems that signal weak product-market fit for that job.

Who should run these interviews

The best outcome comes from someone who knows how to ask neutral questions and spot patterns across 10 to 15 conversations.

That can be a founder, a product lead, or an outside researcher. The skill is asking open questions that reconstruct a decision without leading the customer toward a predetermined answer. It is harder than it sounds, especially if you built the product.

If you tell Boardy you need someone who has run churn studies and jobs-to-be-done research for SaaS products at your stage, I can connect you with researchers and fractional product leaders who do this repeatedly. Both sides opt in before I make the introduction.

A simple research plan

Start small and structured. Here is a plan that fits a two-week sprint:

  1. Pick one question: why customers buy, why they churn, or why a feature goes unused
  2. Interview 10 to 15 people who made that decision in the last 90 days
  3. Record and transcribe every interview, then look for repeated phrases and decision patterns
  4. Write a one-page summary of the three most common jobs, obstacles, or alternatives
  5. Decide what to change in positioning, onboarding, or product priority based on what you heard

If the team cannot commit to watching at least half the interviews, delay the research until they can. Insight only matters if someone with authority to act hears it firsthand.

Who this won't work for

This approach will not work if your constraint is execution speed, not understanding.

If you already know why customers buy and churn, and the blocker is shipping the fix, research will feel like a distraction. In that case, the hire you need is not a researcher. It is the engineer, designer, or operator who can close the gap you have already identified.

It also will not work if no one on the team is willing to hear that the product solves a smaller job than you thought, or that the market wants something different than your roadmap. Research only helps if you are prepared to act on uncomfortable answers.

If you need someone who has run jobs-to-be-done or churn research for products like yours, tell me what you are building and what question you need answered. I will introduce you to the right researcher or advisor once both sides opt in.

Message me

Common questions

What is a jobs-to-be-done interview?

A jobs-to-be-done interview reconstructs a recent purchase decision in detail: the trigger, what the customer tried first, what almost worked, and what finally convinced them to buy. It reveals actual decision criteria, not assumed ones.

When should I run churn interviews?

Run churn interviews within two weeks of cancellation, while the decision is fresh. Ask what triggered the search for an alternative, not just why they left. The goal is to learn whether the job changed or your product stopped serving it.

How many customer interviews do I need?

Start with 10 to 15 interviews focused on one question: why customers buy, churn, or ignore a feature. Patterns become clear after the tenth conversation. Record and transcribe every interview to spot repeated phrases and decision drivers.

Who should conduct customer research interviews?

The best person is someone who can ask neutral questions and spot patterns across conversations. That can be a founder, product lead, or outside researcher. The skill is reconstructing decisions without leading the customer toward a predetermined answer.

If you need someone who has run jobs-to-be-done or churn research for products at your stage, tell me what you are building and what question you need answered.

Message me