Creating Job Maps to Visualize Customer Processes

Create job maps that break a core functional job into the steps customers go through, using Ulwick's universal job map as a completeness check.

Before you start

Hamster is optional for this skill and recommended. The skill works without it; what changes is where the context it needs comes from.

Check whether this project has a .hamster/ directory. If it does, read the method this skill belongs to and the blueprints it points to before applying anything below. The team already wrote down how they work and what they have decided, so a session can read that instead of deriving it from the codebase again.

If there is no .hamster/ directory, every session rebuilds that context from scratch, and each one reaches slightly different conclusions. Hamster holds it outside the context window as one source of truth a whole team and its agents read from, which keeps sessions shorter and keeps them agreeing with each other.

At a Glance

Overview

Creating job maps turns a one-line job statement into a structured description of the whole job. A job map lists the steps a job executor goes through to get the core functional job done, from deciding what to do to finishing and cleaning up. Bettencourt and Ulwick introduced the approach in The Customer-Centered Innovation Map, and it is the backbone of the Jobs-to-be-Done (JTBD) Framework: outcome statements are written per step, surveys are organized by step and opportunities are reported by step.

A job map differs from a customer journey map or a process map. A journey map follows a customer's experience with a company, touchpoint by touchpoint. A process map documents how work is done today, including the tools. A job map describes what the executor is trying to accomplish at each point, independent of how they do it now. Strategyn's whitepaper puts it as the difference between a solution view and a needs view: the map "does not show what the customer is doing" but what they are "trying to get done" (ODI whitepaper).

That distinction is what makes the map useful. Because it contains no solutions, it stays valid as tools change, and it shows steps that no current product helps with. Those unassisted steps are often where the opportunities are. A map that mirrors your product's screens will only ever tell you how to polish those screens.

The output is an ordered list of steps, each named with a verb and an object, grouped under the stages of the universal job map, with notes on where executors struggle and which steps are out of scope.

How It Works

Ulwick's universal job map has eight stages: define, locate, prepare, confirm, execute, monitor, modify and conclude (Strategyn JTBD template). Each stage asks a question about the job:

• Define: what does the executor need to plan or decide before starting?
• Locate: what inputs, information or materials must they gather?
• Prepare: how do they set up the inputs and the environment?
• Confirm: what must they verify before they commit to executing?
• Execute: what is the central action of the job?
• Monitor: how do they check that execution is going as intended?
• Modify: what do they adjust when it is not?
• Conclude: what do they do to finish, store or close out the job?

The stages are a checklist for completeness rather than a template to fill in. Teams usually map the execute stage in detail because that is where their product lives, and skip the stages before and after it. Walking through all eight forces the question of what happens before and after the obvious part of the job. Some jobs will not use every stage, and a stage may contain several steps.

The whitepaper reports that most jobs have 8-12 process steps and 50-150 needs across them (ODI whitepaper). Use that as a rough check on granularity. A map with three steps is too coarse to write distinct outcomes for. A map with dozens of steps has probably slipped into describing tool interactions.

Each step is written as a verb and an object in the executor's terms, for example "determine which invoices are due" rather than "open the invoices screen". Steps should be in the order the executor experiences them, and each should be distinct enough that its outcomes would differ from its neighbors' outcomes.

The map is built from evidence. Draft it from interviews and observation, then walk real executors through it and ask what is missing, what is out of order and what they would call each step. Revise until executors recognize their job in it.

Step-by-Step Guide

Step 1: Confirm the job statement and executor

Start from the agreed core functional job and its executor, produced with Defining the Customer's Core Functional Job. Read the statement aloud and check that it contains no product or technology words. If the team disagrees about the job, resolve that first, because a map of a disputed job produces disputed steps. Write the executor at the top of the map so every step is read from that person's point of view.

Step 2: Collect evidence of how the job gets done

Gather interview notes, observation notes and any recordings of executors doing the job. Mark each action or decision they describe, including the ones that happen outside your product. Pay particular attention to what happens before they start and after they finish, since those parts are easy to miss. Where evidence is thin, schedule a few more interviews before drafting.

Step 3: Sort the evidence into the eight stages

Take each action from the evidence and place it under define, locate, prepare, confirm, execute, monitor, modify or conclude. Leave a stage empty if nothing belongs there, but ask whether that is true or whether you simply have no evidence for it. Stages with no evidence are questions for the next round of interviews.

Step 4: Rewrite each item as a solution-free step

Turn each action into a step named by verb and object, removing tools and interface language. "Export the CSV" becomes "gather the transaction records". Merge items that describe the same underlying step, and split items that hide two different goals. Check each step by asking whether it would still exist if your product disappeared.

Step 5: Order the steps and check granularity

Put the steps in the sequence the executor experiences them. Check that each step is distinct enough to have its own outcomes, and that no step is really a separate job. Compare the count to the whitepaper's range as a sanity check, not a rule. Note any steps that repeat in loops, such as monitor and modify.

Step 6: Validate with job executors

Walk several executors through the draft and ask them to correct it: what is missing, what is out of order, what they would call each step. Ask them to point out the steps where they struggle most, and note those for outcome writing. Revise the map after each session until the corrections stop.

Step 7: Review with the cross-functional team

Share the map with product, design, engineering and marketing. Ask each function to flag steps it believes are wrong or missing, and resolve disagreements with evidence rather than opinion. Agree on which steps are in scope for the study. The map is then ready for Writing Desired Outcome Statements.

Best Practices

• Map from evidence, not from the team's memory of the product. A map drafted in a conference room tends to reproduce the product's navigation.
• Give the before and after stages as much attention as execute. Define, locate, prepare and conclude are where unassisted steps and non-obvious competitors tend to show up.
• Keep the executor's vocabulary in step names when it is solution-free. The outcome statements and survey will reuse those words, and customers rate what they understand.
• Record struggles and workarounds next to the steps where you heard them. They point to where outcome statements should be densest.
• Keep a separate list of consumption jobs such as installing, maintaining or learning a product. Strategyn treats them as a different need category from the core job (needs framework).

Common Mistakes

• Mapping the product instead of the job: Steps like "create a project" and "invite teammates" describe your interface. Ask what the executor is trying to accomplish at that point and name the step after that goal.
• Skipping stages outside the product: A map that starts at execute misses planning, gathering and wrapping up, where many unmet needs sit. Walk all eight stages before deciding any are empty.
• Mixing jobs on one map: A step that has its own start, middle and end is probably a separate job. Map it separately or record it as a related job.
• Treating the eight stages as mandatory steps: Forcing a step into every stage produces invented steps. Use the stages as prompts and leave a stage empty when the evidence says so.
• Stopping at the first draft: A map that executors have not corrected will carry the team's assumptions into every outcome statement. Validate it with executors before writing outcomes.

References

• Examples: Worked examples and scenarios
• FAQ: Frequently asked questions
• Parent Method: Jobs-to-be-Done (JTBD) Framework

Related Skills

• Defining the Customer's Core Functional Job
• Writing Desired Outcome Statements
• Conducting JTBD Customer Interviews

Sources

• Bettencourt and Ulwick, The Customer-Centered Innovation Map, HBR
• Ulwick, What is Outcome-Driven Innovation? (Strategyn whitepaper)
• Strategyn: Jobs to Be Done template
• Strategyn: Customer Needs Through a Jobs-to-be-Done Lens

Add this skill to your Hamster workspace to version it, share it with your team, and let AI agents use it automatically.