---
title: Reverse-Engineering Your ICP From The Data You Already Have
slug: books/agentic-gtm-engineering/reverse-engineering-your-icp-from-the-data-you-already-have
docTags:
createdAt: 2024-12-11T08:40:45.121Z
---
## Step 1: Find Your Winners
First, let's identify your best customers. Not who you think they are. Who they actually are.
Query your CRM for accounts that meet these criteria:
- Highest net revenue retention (over 100%)
- Fastest time to value (implementation under 30 days)
- Lowest support ticket ratio (under industry average)
- Highest product adoption rate (using core features)
- Longest retention (over 12 months)
:::hint{type="info"}
**💪 Practical tip**
Start with this SQL-like thinking 👇
:::
```sql
SELECT
account_name,
revenue_retention,
days_to_implement,
support_tickets_per_month,
feature_adoption_rate,
months_retained
FROM accounts
WHERE
revenue_retention > 100
AND days_to_implement < 30
AND months_retained > 12
ORDER BY revenue_retention DESC
```
## Step 2: Group Your Winners
Now cluster these accounts based on common characteristics:
::::VerticalSplit{layout="middle"}
:::VerticalSplitItem
**Industry signals**
- Company size (revenue, employees)
- Industry/vertical
- Geographic location
- Tech stack (what other tools they use)
- Growth stage
:::
:::VerticalSplitItem
**Behavioral signals**
- How they found you (acquisition channel)
- Their buying process (sales cycle length)
- Decision-maker level (C-suite, VP, Manager)
- Implementation approach (self-serve vs. high-touch)
:::
::::
## Step 3: Find The Patterns
Look for overlaps in these characteristics:
1. Map out the common traits among your top 20% of accounts
2. Identify which traits appear most frequently
3. Calculate success rates for each characteristic
Example pattern analysis:
| **Top revenue segment** | $10M-50M ARR (65% of winners) |
| ----------------------- | ----------------------------------------- |
| **Primary industry** | SaaS (72% of winners) |
| **Common tech stack** | Salesforce + HubSpot (58% of winners) |
| **Typical buyer** | VP of Revenue Operations (45% of winners) |
## Step 4: Validate The Signals
Cross-reference your patterns against:
- Win rates by segment -> look for segments with win rate over 20%
- Customer acquisition costs
- Lifetime value
- Implementation success rates
- Product usage patterns
Create a simple scoring matrix:
## Step 5: Build Your Anti-Portfolio
Just as important... Identify who you shouldn't target:
- Accounts that churned within 6 months
- High support cost ratio
- Low product adoption
- Slow sales cycles
- Poor revenue retention
This helps you spot red flags early.
## Step 6: Turn Insights Into Actions
Create three simple tools from your analysis:
::::::WorkflowBlock
:::::WorkflowBlockItem
### The Ideal Customer Checklist
::::VerticalSplit{layout="middle"}
:::VerticalSplitItem
**Must have:**
Revenue between $10M-50M SaaS business model Using Salesforce + HubSpot Has dedicated RevOps team
:::
:::VerticalSplitItem
**Nice to have:**
Series B+ funding US-based 50+ employees
:::
::::
:::::
:::WorkflowBlockItem
### The Red Flag List
Automatic disqualifiers:
- No dedicated ops team
- Revenue under $5M
- Non-tech industry
- No current marketing automation
:::
:::WorkflowBlockItem
### The Qualification Script
Key questions to ask:
1. What's your current tech stack?
2. Who owns RevOps?
3. What's your implementation timeline?
4. What's your growth target this year?
:::
::::::
## The Real Work
Your ICP should inform:
- Ad targeting parameters
- Content creation focus
- Sales team training
- Product roadmap priorities
- Customer success protocols
:::hint{type="info"}
🧠**Remember**
Your ICP isn't a document you create once and file away. It's a living framework that should get smarter with every customer interaction.
:::
Start with the data you have.
Let it tell you who you should be targeting.
Then build systems to get better at finding them.
That's how you engineer growth.