--- name: journey-map description: "Map and improve any customer experience by the MOMENTS people actually remember, not the average. Distill a journey to its 5-6 key Moments (Peak/Pit/Rise/Dip/Milestone), then improve it with the 4 Peak-End tactics — mark the milestone, elevate the peak, fill only the biggest pit, reorder to end high. Triggers: '/journey-map', 'map the customer journey', 'map the journey', 'improve onboarding', 'onboarding journey', 'peak-end', 'peak end rule', 'where does the experience drop', 'fill the pit', 'elevate the peak', 'end on a high note', 'reorder the flow', 'remembered moments', 'ROI of delight', 'labor illusion'. Self-improving: every run ends with a gotcha review." --- # Journey Map Map an experience down to the 5-6 Moments the brain will actually remember, classify each one, then improve the journey with the four Peak-End tactics. Output is a compact journey map (Moments on a Charge timeline, emotions as faces) plus a ranked, principle-cited set of improvement ideas. The memory of an experience IS the product — this skill designs that memory. This skill is **self-improving**. Every run must end with a gotcha review (see [Self-Improvement Protocol](#self-improvement-protocol)). > **This skill is a checklist, not a suggestion.** Every mode below is a numbered checklist. You MUST tick each `[ ]` by writing its status + exit-criterion outcome in your response BEFORE moving to the next step. Skipping a step without documenting why is a skill violation. --- ## Pre-flight — ALWAYS (before any mode) - [ ] **PF-1. Read [references/gotchas.md](references/gotchas.md).** Exit: you can state the 3 most recent dated entries in one sentence each (or confirm the log holds only the format header). - [ ] **PF-2. Name the ONE journey in scope.** Exit: a single slice of experience with a start and end written down (e.g. "first-run onboarding: install → first success"), NOT the whole product. - [ ] **PF-3. Confirm you can walk the REAL journey.** Exit: you have the actual steps (screens, flows, a recording, or a described walkthrough) — not an imagined ideal. If you only have the ideal, run `user-empathy` to build the Story Strip first, then map the real thing here. - [ ] **PF-4. Read [references/journey-model.md](references/journey-model.md).** Exit: you can name the 5 Moment types and state what the vertical axis is. - [ ] **PF-5. Pick the mode.** Exit: one of Map / Improve / Verify written down. ## Modes | Mode | When to use | Output | |---|---|---| | **Map** | You have an experience but no shared picture of it | A 5-6 Moment journey map on a Charge axis, each Moment classified + given an emotion face | | **Improve** | You have a map and want the experience to feel better | Ranked improvement ideas (biggest impact / lowest effort), each tied to one of the 4 tactics + its principle | | **Verify** | You changed the journey, or want to check the map is honest | Real map compared against the ideal Story Strip; a re-map after changes | --- ## Mode: Map Distill the REAL journey to the handful of Moments the brain will reconstruct it from. Read [references/journey-model.md](references/journey-model.md) for the full mapping discipline. - [ ] **M-1. List the raw steps of the REAL journey.** Exit: an ordered list of what actually happens now — not the ideal, not the happy path only. Include the dropout path if it exists. - [ ] **M-2. Score each step's Charge (the health bar).** Exit: every step marked as roughly high / mid / low Charge (Net Charge = Drive − Friction). Rough is fine — Charge is a mental model, not a spreadsheet. - [ ] **M-3. Distill to 5-6 KEY Moments only (Miller's Law).** Exit: exactly 5-6 Moments survive. An exhaustive map is a map nobody uses. If you have 12 Moments, you have not finished distilling. - [ ] **M-4. Classify each kept Moment.** Exit: each of the 5-6 is labeled **Peak** (highest absolute Charge) · **Pit** (lowest) · **Rise** (a climb) · **Dip** (a fall) · **Milestone** (start/end of a milestone). At minimum the map names the single Peak, the single Pit, and the key Milestone. - [ ] **M-5. Attach an emotion FACE to each Moment.** Exit: every Moment carries a face/emoji for the felt emotion (Pareidolia — faces communicate feeling far better than abstract dots). No bare dots. - [ ] **M-6. Draw it on the Charge axis.** Exit: Moments plotted left-to-right in order with the VERTICAL axis = Charge level, so the peaks and pits are visible as a shape, not a table. - [ ] **M-7. Pass the [Pre-Ship Gate](#pre-ship-gate).** Exit: every G-rule returns PASS. - [ ] **M-8. Update gotchas.** Exit: any new symptom/fix appended to [references/gotchas.md](references/gotchas.md), or a "nothing new" line added. ## Mode: Improve Apply the four tactics that work BECAUSE the brain reconstructs experiences from peaks, pits, and milestones — not averages. Read [references/improvement-tactics.md](references/improvement-tactics.md) for the mechanism behind each. - [ ] **I-1. Confirm a current map exists.** Exit: a 5-6 Moment map (from Map mode) is in hand with Peak, Pit, and Milestone identified. No map → run Map mode first. - [ ] **I-2. Find the root cause of the biggest Dip.** Exit: for the biggest Dip, look at the step slightly BEFORE it — that is usually the cause — and name it. Same for the biggest Rise (so you can amplify what works). - [ ] **I-3. Apply the 4 tactics, in this priority order:** - **Mark the Milestone** — celebrate the milestone start/end so the brain files it as an event. Exit: one concrete way to mark the key Milestone. - **Elevate the Peak** — the brain over-weighs the Peak, so make the highest Moment higher (lean on the user's top drives to do it). Exit: one way to raise the Peak. - **Fill ONLY the biggest Pit** — do NOT chase every small dip. Small pits are potholes; filling them all is whack-a-mole (Pareto: the biggest Pit dominates the memory). Exit: one fix for the single biggest Pit, and an explicit note that the smaller pits are intentionally left alone. - **Reorder to end HIGH** — the End is weighted as heavily as the Peak (Peak-End Rule, Kahneman), so arrange the flow to finish on the strongest Moment; and deliver a PREVIEW of value early, because people prefer a small reward now to a big one later (Hyperbolic Discounting). Exit: one reorder that ends higher and/or previews value sooner. - [ ] **I-4. Consider leverage plays.** Exit: checked whether any unavoidable WAIT can be turned into perceived progress via the Labor Illusion (show the behind-the-scenes work), and whether the ROI-of-delight framing applies — moving good customers (NPS 5-8) to delighted (9-10) is reported at ~9× the revenue of satisfying average ones. **Cite the ~9× as directional, not gospel** (industry study, 2017); use it to justify investing in delight, never as a precise forecast. - [ ] **I-5. Rank the ideas by impact ÷ effort.** Exit: the top 3 ideas listed (biggest impact, lowest effort first); each idea names its tactic AND the underlying principle (e.g. "Reorder — end high, Peak-End Rule"). - [ ] **I-6. ETHICS GATE.** Exit: any idea that adds urgency, scarcity, a countdown, a "labor illusion" that misrepresents real work, or otherwise persuades a shippable surface has been run through the `ethics-check` skill's quick gate (Regret test · scarcity authenticity · defaults-for-the-user). A Labor Illusion that fakes work the system isn't doing FAILS — do not ship it. - [ ] **I-7. Pass the [Pre-Ship Gate](#pre-ship-gate).** Exit: every G-rule returns PASS. - [ ] **I-8. Update gotchas.** Exit: any new symptom/fix appended to [references/gotchas.md](references/gotchas.md), or a "nothing new" line added. ## Mode: Verify Check the map tells the truth, and re-check after changes. - [ ] **V-1. Compare the REAL map against the IDEAL Story Strip.** Exit: the real journey map is set beside the ideal Story Strip from the `user-empathy` skill; every gap between what you WANT the customer to feel and what they ACTUALLY feel is listed. The gaps ARE the improvement backlog. - [ ] **V-2. Sanity-check honesty.** Exit: confirmed the map shows the real journey (including friction and dropout), not a sanitized happy path. A map with no Pit is almost always dishonest — find the Pit. - [ ] **V-3. Re-map after changes.** Exit: if improvements were applied, the journey is re-mapped and the new Peak/Pit/End are confirmed to have moved as intended (especially: does it now END high?). - [ ] **V-4. Pass the [Pre-Ship Gate](#pre-ship-gate).** Exit: every G-rule returns PASS. - [ ] **V-5. Update gotchas.** Exit: any new symptom/fix appended to [references/gotchas.md](references/gotchas.md), or a "nothing new" line added. --- ## Pre-Ship Gate (MUST pass before delivering a map or improvement set) Every map / improvement set about to be delivered must pass every row. Mark each `[ ]` PASS / FAIL. A FAIL blocks delivery — fix it, don't ship the issue. - [ ] **G0. 5-6 Moments, no more.** More than 6 key Moments = not distilled = map won't be used (Miller's Law). FAIL if the map lists 7+. - [ ] **G1. Real, not ideal.** The map reflects what actually happens (friction, dropout included), not the aspirational flow. A Pit-free map FAILS unless proven honest. - [ ] **G2. Vertical axis = Charge, emotions as faces.** FAIL if Moments are plotted on anything but a Charge level, or shown as bare dots instead of emotion faces. - [ ] **G3. Only the biggest Pit is filled.** Improvement sets that try to fill every small pit FAIL (whack-a-mole; Pareto). Smaller pits must be explicitly left alone. - [ ] **G4. It ends HIGH.** Any improvement set FAILS if the resulting journey does not end on a strong Moment (Peak-End Rule) — the End is the most-remembered Moment after the Peak. - [ ] **G5. Every idea names its principle.** No bare opinion. Each recommendation cites its tactic + psychology principle (Peak-End, Hyperbolic Discounting, Labor Illusion, Pareto). FAIL otherwise. - [ ] **G6. Persuasion is gated + numbers are directional.** Any urgency/scarcity/labor-illusion idea passed `ethics-check`; the ~9× ROI-of-Delight is cited as directional, never as a precise promise. FAIL otherwise. --- ## Shared Infrastructure - **The 5 Moment types + mapping discipline:** [references/journey-model.md](references/journey-model.md) - **The 4 improvement tactics + why the brain reconstructs experiences:** [references/improvement-tactics.md](references/improvement-tactics.md) - **Gotchas log:** [references/gotchas.md](references/gotchas.md) --- ## Self-Improvement Protocol Every run must end with a gotcha review. New learnings die silently if you don't log them. **During the run:** - Undocumented error / dead-end → capture symptom + resolution verbatim. - A journey pattern or Moment type that didn't classify cleanly → note the actual shape. - A tactic that misfired or an ROI/behavior surprise → note the measurement. - User pushback corrected your phrasing or methodology → ALWAYS log this. Pushback is the highest-value gotcha source. **At end of run:** 1. **Promote the pattern** to its permanent home BEFORE appending the log entry: - New mapping rule / Moment-classification nuance → [references/journey-model.md](references/journey-model.md) - New improvement tactic / mechanism / ROI nuance → [references/improvement-tactics.md](references/improvement-tactics.md) - New hard rule → Pre-Ship Gate above 2. **Then** append a 1–3 line entry to [references/gotchas.md](references/gotchas.md). Each entry must point to the promoted reference where the full fix lives. 3. **Never** silently let a new learning die. If you discovered it, log it. **Gotcha entry format (1–3 lines):** ```markdown ### YYYY-MM-DD — Symptom → cause → fix. See [#](#) for the full pattern. Context: . ``` Use today's absolute date (the user's current date) — never relative. --- ## Failure / Recovery | Symptom | Cause / Fix | |---|---| | Map has 10+ Moments, nobody uses it | Not distilled. Keep only the single Peak, single Pit, key Milestone + the 2-3 biggest Rises/Dips. See [references/journey-model.md](references/journey-model.md#distill). | | Map has no Pit / looks all-positive | You mapped the ideal, not the real journey. Walk the actual flow incl. dropout; find where Charge bottoms out. G1. | | Filled a pit, experience still feels bad | You filled a pothole, not THE Pit. Refocus on the single lowest Moment (Pareto). See [references/improvement-tactics.md](references/improvement-tactics.md#fill-pit). | | Improvements added but memory unchanged | The End didn't move. Reorder so the journey finishes on the strongest Moment (Peak-End). G4. | | Recommendation sounds like opinion | Missing principle citation. Attach the tactic + named principle. G5. | | Delight idea feels manipulative (fake countdown, faked "working…") | Run `ethics-check` quick gate; a Labor Illusion that misrepresents real work fails the Regret test. I-6 / G6. | | Don't have an ideal to compare against | Run the `user-empathy` skill to build the Story Strip first, then Verify. V-1. | --- ## Related - `user-empathy` — produces the ideal Story Strip (the target feeling) that Map's real journey is compared against in Verify mode; also supplies the top drives used to Elevate the Peak. - `psych-audit` — the Moment-level Imprint Loop / Charge audit; use it to find WHY a specific Dip or Pit happens (this skill locates the Moment, `psych-audit` diagnoses the screen). - `ethics-check` — mandatory gate before shipping any persuasion-flavored improvement (urgency, scarcity, labor illusion). See I-6 / G6. - `../../rules/product-psychology-context.md` — the plugin-wide ethics contract binding these skills together.