What each lane is actually capturing
Actions are the observable behaviours — the things a camera would see. "Searches for a replacement part number", not "realises they need a part". Keeping this lane behavioural is what stops the whole map drifting into speculation, because actions are the row you can most easily check against real data.
Touchpoints are where those actions happen: the search results page, the invoice PDF, the call queue, the box on the doorstep, the colleague who already owns one. A touchpoint is any surface the person and the organisation meet across, including ones you do not control. Thoughts are the questions and judgements running underneath — "is this the right part?", "why do I have to enter this again?" — ideally in the words your research actually captured rather than a tidied-up paraphrase.
Pain points are the friction: the dead ends, the waits, the moments of doubt, the places where people give up or go around you. Opportunities are the only forward-looking lane, and they should read as responses to something specific elsewhere in the same column. An opportunity with no pain point, thought or action above it is usually an idea someone brought into the room, not something the journey surfaced.
The emotion curve, and how to keep it honest
Rate each stage on a simple five-point scale and the tool draws the line above the grid. The value is not precision — nobody believes a journey is exactly a 2.0 at checkout — it is comparison. The curve makes the shape of the experience legible at a glance: where it peaks, where it drops, and how far the drop is from the stage everyone in the organisation spends their time optimising.
The failure mode is rating the stage from the inside. Teams rate "onboarding" a 4 because onboarding was recently redesigned and the team is proud of it, while the research notes underneath say people were confused and anxious. Anchor each rating to something in the Thoughts or Pain Points cells directly below it. If a stage is rated low and the column is empty, either the research is missing or the rating is a feeling about the org, not the customer.
One more thing the curve is good for: it exposes the stages nobody owns. The dips are very often in the gaps between teams — waiting, chasing, being handed off — precisely because no single team's metrics cover them.
Where the content comes from — this is not a desk exercise
A journey map is a synthesis artefact. It is meant to be filled from interviews, support tickets, session recordings, sales call notes, churn surveys, store observations and analytics — evidence about people who are not in the room. Built from memory in a workshop, it reliably reproduces the org chart: the stages become internal process steps, and the pain points become the ones the loudest person in the room already wanted budget to fix.
A workable order is to run the research first, cluster the raw quotes and observations, and only then open the grid and place them. Where you have no evidence for a cell, leave it empty rather than filling it in — an empty cell is a research question, and a map with visible holes is far more useful than a complete-looking one built on assumptions.
It also helps to name the person and the scenario before you start. "A first-time buyer replacing a broken part under time pressure" produces a very different map from "a procurement manager reordering quarterly", and a single map that tries to cover both ends up describing neither.
How it differs from a service blueprint and a user story map
A service blueprint is the same grid extended downwards, past a "line of visibility", into everything the customer never sees: frontstage staff actions, backstage systems, and the support processes underneath. Journey map answers what the experience is like; blueprint answers why it is like that, operationally. If your conversation keeps turning into "the reason that step is slow is the legacy system", you have outgrown the journey map and want a blueprint.
A user story map is a planning artefact, not a research one. Its columns are the activities a user performs in a product, and the cards beneath them are slices of work to be built and released. Journey maps describe the current experience across every channel including the ones you don't build; story maps describe future software in release order. The two are frequently confused because both are grids of cards with a horizontal spine.
There is also the empathy map, which zooms into one person at one moment rather than across time, and the experience map, which is essentially a journey map not tied to a particular company's customers. Pick by the decision you need to make, not by which template looks most impressive.
The failure that wastes the most journey maps
The commonest one is mapping the journey you wish people had. The map follows the funnel as designed — discover, evaluate, sign up, activate, renew — while real people arrive mid-journey from a comparison site, ignore the onboarding, and never touch the feature the whole map is organised around. That map is a picture of the strategy, and it will never surprise anyone, which is the only thing it was worth doing for.
The tells are worth learning: every stage is a step in an internal process, the emotion curve is smooth, no stage sits below neutral, and nothing on the map contradicts what leadership already believes. Real journeys are lumpy, loop backwards, and contain long stretches where nothing happens because the person is waiting or has forgotten about you entirely. Non-linear paths are fine — model the loop as a stage, or note the re-entry point in the Actions lane.
Second commonest: the map gets made, photographed, and never touched again. It is a living document about a moving thing. Keep it where the work happens, revisit it when the product or the market shifts, and let the opportunity lane feed an actual backlog — otherwise it is a poster.
Running the session, and keeping the map alive after
Give people the research to read beforehand, then build the map together rather than presenting a finished one. Start with the stages and argue about them — the boundaries between stages are usually the most productive disagreement in the room — then fill Actions all the way across before touching any other lane, so nobody anchors on a favourite problem. Rate the emotion last, when the evidence is all on the wall.
Everything here saves in your browser and prints to a single sheet for the wall. Saving it to an OmniCanvas canvas keeps the whole grid — stages as columns, lanes as rows, cards in their cells — next to the interview notes, screenshots and tickets it was synthesised from, so the map stays connected to its evidence instead of becoming a standalone picture nobody can audit.