charlieviettq/awesome-agent-skill / .claude/skills/biz-brand-positioning
"biz-brand-positioning"
Verified"Develop brand positioning strategy including positioning statements, perceptual maps, and brand personality/archetype analysis. Use this skill when the user needs to define or refine how their brand is perceived relative to competitors, craft a positioning statement, build a brand identity framework, or map competitive positions — even if they say 'what makes us different', 'our brand feels generic', or 'how do customers see us vs competitors'.".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.0 KB |
| examples/sample_scenario.md | markdown | 6.3 KB |
| references/brand-archetypes.md | markdown | 10.5 KB |
| references/perceptual-mapping.md | markdown | 10.3 KB |
references/perceptual-mapping.md
View on GitHub# Perceptual Mapping — Method Reference
Perceptual mapping translates qualitative brand perception into a 2D visual space. This reference covers axis selection, data collection methods, plotting, and interpretation — specifically for use in the `biz-brand-positioning` skill.
---
## The Core Problem: Axes Determine Everything
A perceptual map is only as useful as its axes. Wrong axes produce a map that is visually neat but strategically useless. The axes must reflect the dimensions customers *actually use* to compare brands — not dimensions that are easy to measure or that make your brand look good.
**Common mistake:** Choosing axes that your brand wins on. If you define the axes, you control where everyone lands. This is a positioning exercise for internal morale, not strategy.
**Correct approach:** Derive axes from customer language.
---
## Step 1: Identify Candidate Dimensions
### Method A — Direct Elicitation (interviews / surveys)
Ask customers: *"When choosing between [Brand X] and [Brand Y], what factors matter most to you?"*
Collect raw language. Do not pre-define categories. Common outputs:
- "feels like a real Taiwanese brand"
- "too expensive for what it is"
- "I trust their quality"
- "it's what my parents drink"
- "for young people"
Group responses into themes. The themes with the highest frequency and the highest stated importance become axis candidates.
### Method B — Repertory Grid Technique
Present customers with three brands at a time (a "triad"). Ask: *"Two of these brands are similar; one is different. In what way are the two similar to each other, but different from the third?"*
Repeat with different triads. The constructs customers generate (e.g., "traditional vs modern", "mass vs craft") are the perceptual dimensions. This method surfaces dimensions customers use implicitly — dimensions they might not name if asked directly.
### Method C — Multidimensional Scaling (MDS)
Collect pairwise similarity ratings for all brands in the competitive set (1–9 scale). Run MDS to produce a low-dimensional map where distance = perceived dissimilarity. The axes in an MDS map are unlabeled — you interpret them post-hoc by examining which attributes correlate with position.
MDS is the most rigorous method but requires statistical software (R: `cmdscale()`, Python: `sklearn.manifold.MDS`) and a large enough sample. Use Method A or B when budget or timeline is limited.
---
## Step 2: Select the Final Two Axes
From your candidate dimensions, select two that meet all three criteria:
| Criterion | Test |
|---|---|
| **Customer-relevant** | Customers mention this dimension unprompted when comparing brands |
| **Discriminating** | Brands in your competitive set actually differ on this dimension |
| **Independent** | The two axes are not highly correlated (e.g., "expensive" and "luxurious" would collapse into one axis) |
If your two best dimensions are correlated (r > 0.7), pick the one that is more actionable or replace one with a different dimension that is uncorrelated.
**Typical axis pairs by category:**
| Category | Axis 1 | Axis 2 |
|---|---|---|
| Consumer packaged goods | Price (low→high) | Naturalness (processed→artisan) |
| F&B / beverage | Brand heritage (new→established) | Target age (young→mature) |
| SaaS / tech | Ease of use (complex→simple) | Feature depth (basic→advanced) |
| Retail | Convenience (inconvenient→convenient) | Exclusivity (mass→premium) |
| Education | Formality (informal→institutional) | Outcome focus (process→results) |
These are starting points only. Derive axes from customer data, not from this table.
---
## Step 3: Score Each Brand
For each brand in the competitive set (including your own), assign a score on each axis.
### Scoring methods (choose one):
**Option A — Perception survey**
Ask a representative sample of your target segment to rate each brand on each axis using a 1–7 Likert scale. Average scores across respondents. This is the most defensible method.
Example survey item:
> "On a scale of 1–7, where 1 = 'Very affordable' and 7 = 'Very premium', how would you rate [Brand]?"
**Option B — Expert judgment panel**
Assemble 3–5 people with deep market knowledge. Each independently rates every brand. Average scores. Flag any brand where ratings diverge by more than 2 points — that brand's position is ambiguous or contested.
**Option C — Proxy data**
For price axes: use actual price data. For distribution axes: use availability data. Proxy data is objective but only works when the axis has a clear real-world analogue.
Do not mix methods across axes in the same map — it produces an inconsistent scale.
---
## Step 4: Plot the Map
### Manual plotting (for client presentations)
Normalize all scores to a 0–10 scale for consistent visual spacing. Place the origin (5, 5) at center. Map the axes:
```
Axis 2 (High)
|
| [Brand C] [Brand A]
|
| [Brand D]
|
| [Brand B] [Our Brand]
|
+---------------------------------- Axis 1 (High)
```
Label each point with the brand name. Do not use logos (they obscure position). Add axis labels at both ends (not just the positive end).
### Python snippet (matplotlib)
```python
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
brands = {
"島嶼啤酒": (8.2, 8.5),
"台灣啤酒": (2.1, 5.0),
"Kirin": (5.0, 2.0),
"Jim & Dad's": (7.8, 5.5),
}
our_brand = "島嶼啤酒"
fig, ax = plt.subplots(figsize=(8, 8))
ax.set_xlim(0, 10)
ax.set_ylim(0, 10)
ax.axhline(5, color="gray", linewidth=0.5, linestyle="--")
ax.axvline(5, color="gray", linewidth=0.5, linestyle="--")
for brand, (x, y) in brands.items():
color = "#D62728" if brand == our_brand else "#1F77B4"
ax.scatter(x, y, s=120, color=color, zorder=5)
ax.annotate(brand, (x, y), textcoords="offset points",
xytext=(8, 4), fontsize=10)
ax.set_xlabel("Price (Low ← → High)", fontsize=11)
ax.set_ylabel("Local Identity (Low ← → High)", fontsize=11)
ax.set_title("Perceptual Map — Taiwanese Beer Market", fontsize=13)
plt.tight_layout()
plt.savefig("perceptual_map.png", dpi=150)
```
---
## Step 5: Interpret the Map
### Identify white space
White space = a quadrant or region with no current occupant, *and* where customers would value a brand positioned there.
**Warning:** Not all white space is opportunity. Ask two questions before calling a gap "strategic white space":
1. Is this quadrant empty because no one wants what it represents? (e.g., "high price + low quality" is always empty — it's not a gap, it's a graveyard)
2. Is this quadrant reachable given your capabilities and cost structure?
White space is only actionable if it combines *unmet demand* with *deliverability*.
### Identify crowding
If 3+ brands cluster in the same region, that space is crowded. A new brand entering there must out-execute incumbents to win — positioning alone won't differentiate. In a crowded map, look at the edges and opposite quadrants.
### Check your own position
If your brand plots near the center of the map, it is perceived as undifferentiated — the worst position. The center means "nothing notable in any direction." This often manifests as customers saying "it's fine, I guess."
---
## Worked Example: Taipei Co-Working Spaces
### Research findings (from 15 customer interviews + 40-person survey)
Target segment: freelancers and small teams (1–5 people) in Taipei.
Top dimensions mentioned: cost, vibe/community, location, flexibility of hours. After correlation check, *cost* and *community vibe* emerged as the two least-correlated, most-discriminating axes.
### Survey results (1–7 scale, n=40)
| Brand | Affordability (1=expensive, 7=cheap) | Community (1=transactional, 7=community-driven) |
|---|---|---|
| WeWork Xinyi | 2.1 | 3.8 |
| Regus | 2.5 | 2.2 |
| The Hive | 3.4 | 5.9 |
| Our brand (草地空間) | 5.1 | 6.2 |
| 獨立咖啡廳 (category anchor) | 6.8 | 4.1 |
### Map interpretation
```
Community (High)
|
| 草地空間 ● ● The Hive
|
| ● 獨立咖啡廳
|
| ● WeWork ● Regus
|
+-----------------------------------------
Cheap Expensive
```
**White space identified:** High community + high price quadrant is empty. No brand currently combines premium pricing with strong community. If the target segment would pay a premium for curated community (requires validation), this is a positioning opportunity.
**草地空間's current position:** Affordable + high community. Differentiated from WeWork/Regus (transactional + expensive) and from coffee shops (affordable but no community infrastructure). The position is defensible but the "affordable" end risks being out-competed on price by coffee shops or by new entrants.
---
## When to Rebuild the Map
A perceptual map reflects perception at a point in time. Rebuild when:
- A major competitor enters or exits the market
- A competitor runs a significant repositioning campaign
- You change your own positioning signals (pricing, distribution, communications)
- More than 18 months have passed since the last research
Do not update the map based on internal assumptions ("we launched a new product, so we've probably moved"). Perception changes slowly and lags execution. Validate with new customer data.
---
## Limitations to Disclose
**Two-dimensional constraint.** Real brand perception is multi-dimensional. A 2D map always compresses information. If two brands score identically on your chosen axes but customers perceive them as very different, your axes are missing something important.
**Average obscures variance.** A brand that polarizes customers (some rate it 2, others rate it 9) will plot in the middle of the map — appearing to be a bland, moderate brand. Always look at score distributions, not just means.
**Perception ≠ preference.** A perceptual map shows where brands are perceived to be, not which position customers prefer. A brand can be distinctly positioned in an undesirable quadrant. You need a separate preference analysis (e.g., ideal-point mapping) to know if a white space is desirable.
**Self-perception bias.** Asking customers to rate your brand on a survey they know you commissioned inflates scores. Use blind labeling ("Brand A, Brand B") when possible for the most accurate data.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |