data-point-research

Data Point Research

You are designing custom data points that predict customer fit, signals that can't be bought from standard data providers but can be researched or inferred.

Core Concept

Build data you can't buy.

Standard data: Employee count, industry, funding stage Custom data: "Has free trial option", "Support team > 20 people", "Uses competitor X"

Input

User provides:

• Their ICP or target segment
• What their product does / pain it solves
• Optionally: existing customer patterns to analyze

Process

1. Identify Predictive Signals

   Prompt Pattern:

   For a company selling [PRODUCT/SERVICE] to [ICP]:
   
   What specific, non-standard data points would strongly predict fit?
   
   Think about:
   - Website indicators (features, pages, content)
   - Hiring patterns (specific roles, urgency)
   - Tech stack combinations (not just "uses Salesforce")
   - Business model indicators (pricing model, customer type)
   - Operational signals (support structure, team ratios)
   

2. Design Detection Methods

   For each data point, define:

   • What to look for
   • Where to find it
   • How to validate it

3. Output Format

   ## Custom Data Points for [ICP/Product]
   
   ### High-Signal Data Points
   
   | Data Point | Why It Predicts Fit | Detection Method |
   |------------|---------------------|------------------|
   | [Point 1] | [Reason] | [How to find] |
   | [Point 2] | [Reason] | [How to find] |
   
   ### Detection Playbooks
   
   #### Data Point: [Name]
   
   **What to Look For:**
   - [Specific indicator 1]
   - [Specific indicator 2]
   
   **Where to Find It:**
   - [Source 1]: [What to check]
   - [Source 2]: [What to check]
   
   **Claygent/AI Prompt:**
   

   Visit [URL pattern]. Look for [specific element]. Return: [Yes/No/Value] based on [criteria]. If not found, return "Not found".

   
   **Validation:**
   - Manual spot-check: [What to verify]
   - False positive risk: [High/Medium/Low]
   
   ### Scoring Model
   
   | Data Point | Weight | Reason |
   |------------|--------|--------|
   | [Point 1] | [+X] | [Why weighted this way] |
   | [Point 2] | [+X] | [Why weighted this way] |
   
   Total threshold for "High Fit": [Score]
   

Examples

Input: "We sell AI customer service software. Best customers have high support volume and are PLG."

Output:

## Custom Data Points for AI Support Software

### High-Signal Data Points

| Data Point | Why It Predicts Fit | Detection Method |
|------------|---------------------|------------------|
| Has free trial/freemium | Indicates PLG model, high volume | Website pricing page |
| Support team size > 20 | High support volume | LinkedIn search, job postings |
| CS job postings | Active investment in support | Job board scraping |
| Login button on homepage | Has end-user product | Website scan |
| Multiple pricing tiers | PLG with expansion potential | Pricing page analysis |
| "Help center" or "Support" prominent | Support is key function | Website structure |

### Detection Playbooks

#### Data Point: Has Free Trial / PLG Model

**What to Look For:**
- "Free trial" or "Try free" CTA
- "Freemium" mentioned
- Self-serve signup flow
- Credit card not required messaging

**Where to Find It:**
- Homepage: CTAs and hero section
- Pricing page: Tier structure
- Sign-up flow: Self-serve vs "Contact sales" only

**Claygent/AI Prompt:**

Visit {{Company Website}}/pricing (or homepage if no pricing page). Does this company offer a free trial or freemium tier? Look for: "free trial", "try free", "freemium", "$0", "no credit card" Return: "Yes - [type]" if found, "No - Sales-led only" if contact sales only, "Unknown" if unclear.

**Validation:**
- Manual spot-check: Visit 10 companies, verify accuracy
- False positive risk: Low (clear signals)

#### Data Point: Support Team Size > 20

**What to Look For:**
- Number of people with "Support", "Customer Success", "Help Desk" in title
- Support-related job postings volume

**Where to Find It:**
- LinkedIn: Company page → People → Filter by title keywords
- Job boards: Search "[Company] support" or "customer success"

**Claygent/AI Prompt:**

Search LinkedIn for [Company] employees with titles containing: "Support", "Customer Success", "Help Desk", "Customer Service" Return approximate count: "<10", "10-20", "20-50", "50+" If unable to determine, return "Unknown".

**Validation:**
- Cross-reference with company size (ratio check)
- False positive risk: Medium (title variations)

### Scoring Model

| Data Point | Weight | Reason |
|------------|--------|--------|
| Free trial/PLG | +30 | Core ICP indicator |
| Support team 20+ | +25 | Volume signal |
| CS job postings | +20 | Active investment |
| Login on homepage | +15 | End-user product |
| Multiple pricing tiers | +10 | PLG indicator |

Total threshold for "High Fit": 60+

Key Principles

1. Specific > Generic: "Has free trial" beats "is SaaS"

2. Detectable > Theoretical: If you can't find it reliably, it's not useful

3. Predictive > Descriptive: Focus on what correlates with becoming a customer

4. Validated > Assumed: Test with actual customer data

Related Skills

• /pain-qualified-segment - Use data points to define segments
• /company-goals - Job postings as data source
• /saas-identification - One type of custom classification
• /ideal-customer-profiles - Connect data points to ICP

Credits

"Build data you can't buy" is Petra Hajal's line and approach. The prompts and wording here are ours.

---

Examples are illustrative. Company names, prices and numbers in them are placeholders or may be out of date, so check the live source before you rely on any detail.