Skip to content

Commit 92cbada

Browse files
committed
docs: add comprehensive usage guide
- Philosophy (evidence, quantified, actionable) - When to use each mode with examples - Effective query patterns - Interpreting results (confidence, signals, red flags) - Follow-up strategies - Common mistakes to avoid - Advanced research patterns - Comparison with /last30days
1 parent 09110cf commit 92cbada

1 file changed

Lines changed: 309 additions & 0 deletions

File tree

GUIDE.md

Lines changed: 309 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,309 @@
1+
# /intel — Complete Usage Guide
2+
3+
A comprehensive guide to using /intel for market research and competitive analysis.
4+
5+
## Table of Contents
6+
1. [Philosophy](#philosophy)
7+
2. [When to Use Each Mode](#when-to-use-each-mode)
8+
3. [Effective Queries](#effective-queries)
9+
4. [Interpreting Results](#interpreting-results)
10+
5. [Follow-up Strategies](#follow-up-strategies)
11+
6. [Common Mistakes](#common-mistakes)
12+
7. [Advanced Patterns](#advanced-patterns)
13+
14+
---
15+
16+
## Philosophy
17+
18+
/intel is built on three principles:
19+
20+
### 1. Evidence Over Opinion
21+
Every claim must have a source. If Claude can't find evidence, it says "low confidence" rather than making things up. This means you can trust the findings.
22+
23+
### 2. Quantified Insights
24+
"Several people mentioned this" is useless. "12 Reddit posts with 847 total upvotes mentioned this" is actionable. /intel always tries to quantify.
25+
26+
### 3. Actionable Output
27+
The goal isn't to inform — it's to help you decide. Every output ends with a recommendation or clear next step.
28+
29+
---
30+
31+
## When to Use Each Mode
32+
33+
### `/intel gaps [topic]`
34+
**Use when:** You're looking for a product idea or feature opportunity.
35+
36+
**Best for:**
37+
- Exploring a market you're curious about
38+
- Finding underserved niches
39+
- Validating that a problem actually exists
40+
41+
**Example queries:**
42+
```
43+
/intel gaps AI video editing
44+
/intel gaps personal finance for freelancers
45+
/intel gaps developer productivity tools
46+
```
47+
48+
### `/intel competitors [product]`
49+
**Use when:** You're entering a market or positioning against existing players.
50+
51+
**Best for:**
52+
- Understanding who you're competing with
53+
- Finding positioning opportunities
54+
- Learning from competitor weaknesses
55+
56+
**Example queries:**
57+
```
58+
/intel competitors Notion
59+
/intel competitors Stripe
60+
/intel competitors my-startup-name
61+
```
62+
63+
### `/intel sentiment [topic]`
64+
**Use when:** You want to understand how people feel about something.
65+
66+
**Best for:**
67+
- Evaluating a tool before adopting it
68+
- Understanding brand perception
69+
- Tracking sentiment changes over time
70+
71+
**Example queries:**
72+
```
73+
/intel sentiment Claude Code
74+
/intel sentiment React vs Vue
75+
/intel sentiment remote work 2026
76+
```
77+
78+
### `/intel pricing [category]`
79+
**Use when:** You're setting prices or researching pricing models.
80+
81+
**Best for:**
82+
- Setting your own pricing
83+
- Understanding what the market will bear
84+
- Finding pricing model inspiration
85+
86+
**Example queries:**
87+
```
88+
/intel pricing email marketing SaaS
89+
/intel pricing developer tools
90+
/intel pricing AI writing assistants
91+
```
92+
93+
### `/intel trends [topic]`
94+
**Use when:** You want to know if something is growing, declining, or stable.
95+
96+
**Best for:**
97+
- Timing market entry
98+
- Avoiding dying technologies
99+
- Spotting emerging opportunities
100+
101+
**Example queries:**
102+
```
103+
/intel trends no-code 2026
104+
/intel trends AI agents
105+
/intel trends Web3 developer interest
106+
```
107+
108+
### `/intel pain [category]`
109+
**Use when:** You want to find problems worth solving.
110+
111+
**Best for:**
112+
- Finding the most acute problems
113+
- Understanding workarounds people use
114+
- Estimating willingness to pay
115+
116+
**Example queries:**
117+
```
118+
/intel pain project management
119+
/intel pain code review
120+
/intel pain hiring developers
121+
```
122+
123+
### `/intel full [topic]`
124+
**Use when:** You want complete market intelligence.
125+
126+
**Best for:**
127+
- Initial market exploration
128+
- Investment research
129+
- Comprehensive due diligence
130+
131+
**Example queries:**
132+
```
133+
/intel full personal finance apps
134+
/intel full AI coding assistants
135+
/intel full sustainable fashion marketplace
136+
```
137+
138+
---
139+
140+
## Effective Queries
141+
142+
### Be Specific
143+
`/intel gaps apps` — Too broad
144+
`/intel gaps task management for remote teams` — Specific audience + category
145+
146+
### Use Natural Language
147+
`/intel competitors productivity-software-2026` — Robotic
148+
`/intel competitors Notion` — Natural
149+
150+
### Include Context When Helpful
151+
`/intel pricing SaaS` — What kind?
152+
`/intel pricing email marketing SaaS` — Clear category
153+
154+
### Match Mode to Intent
155+
`/intel gaps Notion` — Gaps is for markets, not products
156+
`/intel competitors Notion` — Competitors for specific products
157+
158+
---
159+
160+
## Interpreting Results
161+
162+
### Confidence Levels
163+
164+
**HIGH Confidence:**
165+
- 15+ sources analyzed
166+
- Findings appear in 3+ sources
167+
- Clear engagement signals (upvotes, likes)
168+
169+
**MEDIUM Confidence:**
170+
- 8-15 sources analyzed
171+
- Some corroborating evidence
172+
- Mixed engagement signals
173+
174+
**LOW Confidence:**
175+
- <8 sources analyzed
176+
- Single-source findings
177+
- Niche topic with sparse discussion
178+
179+
### Signal Strength
180+
181+
Not all mentions are equal:
182+
183+
| Signal | Weight | Why |
184+
|--------|--------|-----|
185+
| Reddit post with 500+ upvotes | High | Community validated |
186+
| HN front page discussion | High | Tech community signal |
187+
| Single Twitter mention | Low | Could be noise |
188+
| Company blog post | Medium | Potential bias |
189+
| Multiple independent mentions | Highest | Pattern confirmed |
190+
191+
### Red Flags
192+
193+
Watch for these in results:
194+
- "Only found 3 sources" — Sparse data, validate elsewhere
195+
- "Contradicting opinions" — Market is fragmented
196+
- "Data from 2024" — Might be outdated
197+
- "Company website says..." — Biased source
198+
199+
---
200+
201+
## Follow-up Strategies
202+
203+
After getting initial results, you can:
204+
205+
### Dive Deeper
206+
```
207+
"Tell me more about finding #2 — the brand consistency gap"
208+
"What specific features do people want for audio-visual sync?"
209+
"Can you find more quotes about the pricing complaints?"
210+
```
211+
212+
### Pivot Direction
213+
```
214+
"Now do /intel pricing for this market"
215+
"What about /intel sentiment on the top competitor you found?"
216+
"Run /intel pain for the gap you identified"
217+
```
218+
219+
### Validate Findings
220+
```
221+
"Are there any contradicting views on this?"
222+
"How recent is this data?"
223+
"What's the sample size on the sentiment analysis?"
224+
```
225+
226+
---
227+
228+
## Common Mistakes
229+
230+
### 1. Taking Results as Gospel
231+
/intel searches public discussions. It doesn't know about:
232+
- Enterprise customers (who don't post on Reddit)
233+
- Silent majority (happy users don't complain)
234+
- Recent changes (last few days)
235+
236+
**Fix:** Use /intel as a starting point, not final answer.
237+
238+
### 2. Ignoring Low Confidence
239+
When /intel says "LOW confidence," it means the data is sparse. Don't build a business on thin evidence.
240+
241+
**Fix:** Run multiple queries, do primary research for validation.
242+
243+
### 3. Searching Too Broadly
244+
`/intel full productivity` returns generic results. The narrower your query, the more actionable the output.
245+
246+
**Fix:** Add specifics — audience, use case, geography.
247+
248+
### 4. Not Following Up
249+
The first result is often just the beginning. The best insights come from drilling into specific findings.
250+
251+
**Fix:** Ask follow-up questions about surprising or important findings.
252+
253+
---
254+
255+
## Advanced Patterns
256+
257+
### Market Entry Research
258+
Run in sequence:
259+
1. `/intel full [market]` — Get the lay of the land
260+
2. `/intel gaps [market]` — Find opportunities
261+
3. `/intel competitors [top competitor]` — Study the leader
262+
4. `/intel pricing [market]` — Understand economics
263+
264+
### Competitive Positioning
265+
1. `/intel competitors [your product]` — See the landscape
266+
2. `/intel sentiment [competitor 1]` — Find their weakness
267+
3. `/intel sentiment [competitor 2]` — Compare
268+
4. Position against their weaknesses
269+
270+
### Problem Validation
271+
1. `/intel pain [category]` — Find top problems
272+
2. `/intel gaps [specific problem]` — Check if solutions exist
273+
3. `/intel sentiment [existing solutions]` — See if they're loved/hated
274+
4. Decide if worth building
275+
276+
### Trend Timing
277+
1. `/intel trends [technology]` — Check direction
278+
2. `/intel sentiment [technology]` — Check current perception
279+
3. `/intel gaps [technology]` — Check unmet needs
280+
4. Decide timing of entry
281+
282+
---
283+
284+
## Comparison with /last30days
285+
286+
| Aspect | /last30days | /intel |
287+
|--------|-------------|--------|
288+
| **Focus** | Prompts & techniques | Business decisions |
289+
| **Audience** | Creators, prompters | Founders, builders |
290+
| **Output** | Copy-paste prompts | Actionable intelligence |
291+
| **Modes** | Single | 7 specialized modes |
292+
| **Best for** | "How do I prompt X?" | "Should I build X?" |
293+
294+
Use /last30days when you need to learn techniques.
295+
Use /intel when you need to make business decisions.
296+
297+
---
298+
299+
## Getting Help
300+
301+
If results are unexpected:
302+
1. Try rephrasing the query
303+
2. Use a different mode
304+
3. Check if the topic is too niche
305+
4. Ask follow-up questions for clarification
306+
307+
---
308+
309+
**/intel** — Stop guessing. Start knowing.

0 commit comments

Comments
 (0)