Replace 7-component template with 8-question scenario dialog

Restructure Phase 3 scenario creation around a natural conversation
flow instead of a template checklist:

- 8 strategic questions asked in logical order: transaction → business
  goal → user & situation → driving forces → device → starting point
  → best outcome → shortest path
- Two modes: conversation (agent asks, user answers) and suggest
  (agent answers from Trigger Map, user reviews)
- Broader transaction definition: includes content browsing and
  research, not just purchases/bookings
- Agent creates first page specification after Q8, giving Phase 4
  a concrete starting point
- Template reordered to match Q1-Q8 dialog flow

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Mårten Angner 2026-02-25 09:31:09 +01:00
parent 5577f1fe06
commit 5e6dd2c973
2 changed files with 132 additions and 89 deletions

View File

@ -1,7 +1,7 @@
# Scenario Outline Template
**Used by:** step-05-outline-scenario.md
**Purpose:** Complete template for one scenario outline with all 7 required components.
**Purpose:** Structure the answers from the 8-question scenario dialog into a complete scenario outline.
---
@ -16,49 +16,55 @@
---
## Core Feature
## Transaction (Q1)
**What this scenario covers:**
[State as user purpose, not feature name]
**Business Goal Alignment:**
[Which specific business goal this serves, with objective reference]
[The key transaction — stated as user purpose, not feature name]
---
## Entry Point
## Business Goal (Q2)
**Goal:** [Which specific business goal this serves]
**Objective:** [Objective reference from Trigger Map]
---
## User & Situation (Q3)
**Persona:** [Name] ([Priority level: Primary/Secondary/Tertiary])
**Situation:** [Real-life context — who they are, where they are, what's happening]
---
## Driving Forces (Q4)
**Hope:** [What they're hoping to find or achieve — one sentence]
**Worry:** [What they're afraid of or want to avoid — one sentence]
> CONSTRAINT: One sentence per component. Phrases, not paragraphs.
---
## Device & Starting Point (Q5 + Q6)
**Device:** [Mobile / Desktop / Tablet]
**Context:** [Where the user is, what they're doing, what just happened]
**Discovery:** [How they found the site — Google search terms, link from email, bookmark, ad, etc.]
**Specific entry:** [Device] + [Context] + [Discovery] — max 2 sentences
**Entry:** [How they actually arrive] — max 2 sentences
---
## Mental State
**Trigger:** [What just happened that brought them here NOW]
**Hope:** [What they're hoping to find or achieve — be specific]
**Worry:** [What they're afraid of or want to avoid — be specific]
> CONSTRAINT: One sentence per component. Use phrases, not paragraphs.
---
## Success Goals
## Best Outcome (Q7)
**User Success:**
[Tangible, measurable outcome the user achieves when scenario completes]
[Tangible, measurable outcome the user achieves]
**Business Success:**
[Specific, measurable result the business gets when scenario completes]
[Specific, measurable result the business gets]
---
## Shortest Path
## Shortest Path (Q8)
[Linear sunshine path — NO branches, NO "if" statements. Minimum viable steps.]
@ -70,7 +76,7 @@
## Trigger Map Connections
**Persona:** [Name] ([Priority level: Primary/Secondary/Tertiary])
**Persona:** [Name] ([Priority level])
**Driving Forces Addressed:**
- ✅ **Want:** [Specific positive driver from Trigger Map]
@ -80,6 +86,15 @@
---
## First Page (Entry Point)
**Page:** [Page name from step 1 of Shortest Path]
**Entry context:** [Device] arriving via [discovery method], feeling [mental state summary]
**What the user needs here:** [Purpose from step 1 of Shortest Path]
**Success criteria:** [What must be true before user moves to step 2]
---
## Pages in This Scenario
| # | Page | Purpose in Flow | Template? |
@ -95,16 +110,16 @@
When filling this template, check:
**Entry Point** — Would a real person arrive this way? Include device, location, search terms.
**Transaction** — Is this a real user journey? Browsing content page-by-page counts. Comparing options counts. Any meaningful path through the site with intent.
**Mental State** — Can you FEEL the user's emotional state? "Interested" is not enough. "Panicked because family vacation is at risk" is.
**Driving Forces** — Can you FEEL the user's state? "Interested" is not enough. "Panicked because family vacation is at risk" is.
**Success Goals** — "Get more customers" fails. "Reduce info calls by 40% by giving tourists the info they need online" passes.
**Best Outcome** — "Get more customers" fails. "Reduce info calls by 40% by giving tourists the info they need online" passes.
**Shortest Path** — Count the steps. Can you remove any? Each step must justify its existence. If you can't defend why a step is there, remove it.
**Shortest Path** — Count the steps. Can you remove any? Each step must justify its existence.
**Trigger Map** — Don't invent a user. Use the actual persona from the Trigger Map with their actual driving forces.
**Trigger Map** — Don't invent a user. Use the actual persona with their actual driving forces.
---
_Template for Step 05: Outline Scenario_
_Template for Step 05: Outline Scenario (8-Question Dialog)_

View File

@ -13,7 +13,7 @@ scenarioTemplate: '../data/scenario-outline-template.md'
## STEP GOAL:
Create a detailed outline for ONE scenario using all 7 required components, verify against quality gates, create the output file, then loop for each remaining scenario in the approved plan.
Define ONE scenario through 8 strategic questions in natural conversation order. Start with the primary transaction (highest priority), complete it fully, then loop for each remaining scenario. A **transaction** is any meaningful user journey — purchasing, booking, researching content page-by-page, comparing options, or any interaction where the user moves through the site with intent.
## MANDATORY EXECUTION RULES (READ FIRST):
@ -35,18 +35,18 @@ Create a detailed outline for ONE scenario using all 7 required components, veri
### Step-Specific Rules:
- 🎯 Focus on ONE scenario at a time, complete it fully before moving to the next
- 🚫 FORBIDDEN to skip any of the 7 required components
- 💬 Approach: Collaborate with user on each scenario, using the template as structure
- 🎯 Focus on ONE transaction at a time, complete it fully before moving to the next
- 🚫 FORBIDDEN to skip any of the 8 strategic questions
- 💬 Approach: Ask one question at a time, let the answer shape the next question naturally
- 📋 Verify all quality gates before proceeding to the next scenario or step
## EXECUTION PROTOCOLS:
- 📖 Load the scenario outline template before starting
- 📋 Ensure all 7 components are present and high quality
- 💬 Walk through 8 questions as a dialog — one question at a time, building on each answer
- ✅ Run quality gates check before moving on
- 💾 Create output file in the correct folder structure
- 🔄 Loop back for each remaining scenario
- 🔄 Loop back for each remaining scenario (next transaction, next target group)
- 🚫 FORBIDDEN to proceed if any quality gate fails
## CONTEXT BOUNDARIES:
@ -69,69 +69,79 @@ If returning from a loop, continue with the next unfinished scenario.
Load the full template: `{scenarioTemplate}`
### 3. Create Outline with 7 Required Components
### 3. The 8-Question Scenario Dialog
Every scenario outline MUST have all 7 components:
**Two modes — same 8 questions, different driver:**
#### Component 1: Scenario Name & ID
- **Conversation mode** (default): YOU ask, the USER answers. One question at a time. Each answer shapes the next question naturally.
- **Suggest mode** (when user asks you to suggest): YOU answer all 8 questions based on the Trigger Map, Product Brief, and VTCs. Present the complete scenario to the user for review and adjustment.
- **Name:** Uses persona name + purpose (e.g., "Hasse's Emergency Search")
- **ID:** 01, 02, etc.
- **Slug:** `01-hasses-emergency-search`
This IS the scenario — when all 8 are answered, the outline writes itself.
#### Component 2: Core Feature
> **What counts as a transaction:** Not just purchases or bookings. Clicking through a menu item by item to research site content is a transaction. Comparing options is a transaction. Any meaningful journey where the user moves through the site with intent.
What this scenario covers, stated as user purpose (not feature name).
#### Q1: "What transaction do we need to get really right?"
Start with the WHY. What is the most important thing a user needs to accomplish on this site?
- State as user purpose, not feature name
- **Bad:** "Homepage and service pages"
- **Good:** "Verify service availability before booking"
Must align with a specific business goal from the Trigger Map.
#### Q2: "If this transaction succeeds, which business goal does it add value to?"
#### Component 3: Entry Point (Realistic)
Connect to the Trigger Map immediately. Which specific business goal and objective does this serve?
How the user ACTUALLY arrives. Be specific about:
- **Device:** Mobile, desktop, tablet
- **Context:** Where they are, what they are doing
- **Discovery:** How they found the site (Google search, link, ad, bookmark)
- Reference actual goals from the Trigger Map
- This grounds the scenario in business strategy, not just user needs
- **Bad:** "User opens app"
- **Good:** "Googles 'car repair Oland' on mobile while parked at gas station, clicks top organic result"
#### Q3: "Which user experiences this most, and in what real-life situation?"
**Length Rule:** Entry Point must be 1-2 sentences max. State device + context + discovery method. No backstory or narrative prose.
Identify the persona AND their context. Not just "who" but "who, where, when."
#### Component 4: Mental State (Trigger / Hope / Worry)
- Use actual personas from the Trigger Map
- **Bad:** "A customer looking for information"
- **Good:** "Hasse, 55, motorhome tourist stranded in Byxelkrok with a broken vehicle during family vacation"
Three components, all required, all specific:
#### Q4: "What do they want and what do they fear going into this interaction?"
- **Trigger:** What just happened that brought them here NOW
- **Hope:** What they are hoping to find or achieve
- **Worry:** What they are afraid of or want to avoid
The driving forces — hope and worry. These must be visceral and specific.
- **Hope:** What they're hoping to find or achieve
- **Worry:** What they're afraid of or want to avoid
- **Bad:** "User is interested in the product"
- **Good:** "Trigger: Motorhome broke down in Byxelkrok, family vacation at risk. Hope: Find trustworthy mechanic nearby, get back on road today. Worry: Being stranded for days, getting ripped off by unknown mechanic"
- **Good:** "Hope: Find trustworthy mechanic nearby, get back on road today. Worry: Being stranded for days, getting ripped off by unknown mechanic"
- **Length Rule:** ONE sentence max per component. Phrases, not paragraphs.
**Length Rule:** Each Trigger/Hope/Worry must be ONE sentence max. Use comma-separated phrases, not paragraphs.
#### Q5: "What device are they on?"
#### Component 5: Success Goals (Mutual Value)
Mobile, desktop, or tablet. This shapes the entire design approach.
Both required, both specific and measurable:
#### Q6: "What's the natural starting point — how do they actually arrive?"
How the user ACTUALLY gets to the site. Be specific about discovery method.
- **Bad:** "User opens the website"
- **Good:** "Googles 'car repair Öland' on mobile while parked at gas station, clicks top organic result"
- **Length Rule:** 1-2 sentences max. Device + context + discovery method.
#### Q7: "What does the best possible outcome look like — for both sides?"
Mutual success — user AND business. Both specific and measurable.
- **User Success:** Tangible outcome the user achieves
- **Business Success:** Measurable result for the business
- **Bad:** User: "Successfully use the site" / Business: "Get more customers"
- **Good:** User: "Confirmed mechanic fixes motorhomes, has location and hours, feels confident calling" / Business: "High-intent tourist call captured, positioned as emergency-capable, info call avoided"
#### Component 6: Shortest Path (Linear Sunshine Path)
#### Q8: "What's the shortest path through the site to get there?"
Numbered steps. Each step has page name + what user accomplishes there.
The linear sunshine path. Numbered steps, each with page name + what the user accomplishes.
**Rules:**
- Completely linear — ZERO "if" statements, ZERO branches
- Minimum viable steps — can you remove any step without breaking the flow?
- Each step moves meaningfully toward success
- This is the happiest path when everything works perfectly
**Format:**
```
@ -140,42 +150,57 @@ Numbered steps. Each step has page name + what user accomplishes there.
3. **[Page Name]** — [What user sees/does/achieves here] ✓
```
#### Component 7: Trigger Map Connections
### 4. Name the Scenario
Explicitly link to Trigger Map data:
- **Persona:** [Name] ([Priority level])
- **Driving Forces Addressed:**
- Positive: [specific want from Trigger Map]
- Negative: [specific fear from Trigger Map]
- **Business Goal:** [specific goal + objective number]
After the 8 questions, name the scenario using the persona:
### 4. Quality Gates (Check Before Moving On)
- **Name:** Persona name + purpose (e.g., "Hasse's Emergency Search")
- **ID:** 01, 02, etc.
- **Slug:** `01-hasses-emergency-search`
### 5. Create the First Page
We now know the natural starting point (Q6) and what the user needs to accomplish there (Q8, step 1). Create the first page specification:
1. Take step 1 from the Shortest Path (Q8)
2. Create: `{output_folder}/C-UX-Scenarios/[NN-slug]/pages/[page-slug].md`
3. Include:
- **Page name** from the shortest path
- **Entry context:** Device (Q5) + how they arrived (Q6) + mental state (Q4)
- **What the user needs here:** The purpose from step 1 of the shortest path
- **Success criteria:** What must be true before the user moves to step 2
This gives Phase 4 (UX Design) a concrete starting point instead of an abstract scenario document.
### 6. Quality Gates (Check Before Moving On)
Before proceeding to the next scenario, verify:
- [ ] All 8 questions answered with specific, concrete responses
- [ ] Mental state is visceral and specific (not generic "interested")
- [ ] Entry point is realistic with device + context + discovery method
- [ ] Path is truly linear (zero "if" statements)
- [ ] Both successes are specific and measurable (not vague)
- [ ] Scenario name includes persona name
- [ ] All 7 components present
- [ ] Trigger Map connection is explicit (persona + business goal)
- [ ] First page specification created with entry context
**If any gate fails:** Fix before proceeding.
### 5. Create the File
### 7. Create the Scenario File
1. Create folder: `{output_folder}/C-UX-Scenarios/[NN-slug]/`
2. Create file: `{output_folder}/C-UX-Scenarios/[NN-slug]/[NN-slug].md`
3. Use the template from data/ to structure the content
3. Use the template from data/ to structure the content from the 8 answers
### 6. Loop Check
### 8. Loop Check
**Are there more scenarios in the approved plan?**
**Are there more transactions in the approved plan?**
- **Yes** → Loop back to instruction 1 for the next scenario
- **Yes** → Loop back to instruction 1 for the next transaction and target group
- **No** → Proceed to menu options
### 7. Present MENU OPTIONS
### 9. Present MENU OPTIONS
Display: "Are you ready to [C] Continue to Generating the Overview?"
@ -200,16 +225,19 @@ ONLY WHEN [C continue option] is selected and [all scenarios from approved plan
### ✅ SUCCESS:
- Each scenario has all 7 required components
- All 8 questions answered for each scenario with specific, concrete responses
- All quality gates pass for every scenario
- Output files created in correct folder structure
- Scenarios processed in priority order
- Scenarios processed in priority order (primary transaction first, then secondary, etc.)
- All scenarios from approved plan completed before proceeding
- Menu presented and user input handled correctly
- Conversation mode: Dialog felt like a natural conversation, not a form to fill
- Suggest mode: All 8 answers grounded in actual Trigger Map/Brief data, presented for user review
### ❌ SYSTEM FAILURE:
- Missing any of the 7 required components
- Skipping any of the 8 strategic questions
- Conversation mode: Presenting all questions at once instead of one at a time
- Suggest mode: Not presenting answers for user review before proceeding
- Proceeding with failing quality gates
- Skipping scenarios from the approved plan
- Using generic mental states or vague success goals