Revenue Workflow

ICP Model Build and Backtest

Interview your best AEs and CSMs, enrich 12+ months of closed deals with firmographic and technographic data, then build and backtest a tiering model that correlates to win rate and contract size.

AdvancedTime to value · 4 to 6 weeks5 tools

The outcome

A validated ICP scoring model that routes acquisition, SDR effort, and ABM spend to accounts most likely to close fast and large.

How it flows

How it flows.

Source

CRM

Process

Closed Won

Process

Closed Lost

Process

Interview AEs and CSMs

Process

Analyze highest spend customers

Process

Find firmographic data

Process

Find technographic data

Process

Look for commonalities

Output

ICP Report

Agent

Scoring Model

Output

Tier 1 Accounts

Output

Tier 2 Accounts

Output

Tier 3 Accounts

Process

Backtest model

Process

Backtest model

Drag to pan · scroll to zoom

Build it, step by step

Build it, step by step.

01
Interview AEs and CSMs
Run a structured survey with top performers to capture who buys fast, who churns, and where deals stall. Use an LLM to cluster answers into recurring patterns.
- Survey 5 to 10 AEs and CSMs who close the most revenue or retain the highest NRR.
- Ask who champions deals, who ghosts, which industries stall, and which segments churn.
- Feed raw responses into Claude or GPT to extract common traits of best-fit and worst-fit accounts.
- Output a qualitative list of positive and negative ICP signals.
02
Export and enrich historical deals
Pull 12 to 24 months of Closed Won and Closed Lost opportunities from your CRM, then enrich each account in Clay with firmographic, technographic, and growth signals.
- Export every closed opportunity (won and lost) with account domain, ACV, close date, and stage history.
- Push the account list into Clay and append firmographics: industry, employee count, funding stage, headquarters region.
- Add technographics: CRM platform, data stack, competitor tools, integrations, and job postings.
- Optionally layer in growth signals like hiring velocity, web traffic, and G2 review momentum.
- Output a single enriched dataset that explains why certain accounts closed and others did not.
03
Define tier criteria and scoring rubric
Translate qualitative signals and enriched data into a three-tier rubric (Tier 1, 2, 3) based on firmographic fit, technographic maturity, and negative indicators.
- Define Tier 1 as high-fit, high-value accounts worth manual prospecting and ABM investment.
- Set firmographic thresholds: sub-industry, headcount range, funding stage, region, and growth stage.
- Add technographic criteria: presence of complementary tools, stack maturity, and absence of conflicting platforms.
- List disqualifiers: wrong industry, team too small, no budget signal, or incumbent competitor deeply embedded.
- Document the rubric in a shareable format (spreadsheet or Notion page) so every GTM team member can reference it.
04
Backtest the model in Clay
Implement the scoring rubric as Clay formulas, run it across every historical deal, and validate that Tier 1 accounts correlate with higher win rates and larger ACV.
- Build the tier logic in Clay using conditional formulas or enrichment waterfall rules.
- Score every past Closed Won and Closed Lost account and assign a tier.
- Pivot the results: group by tier and calculate average win rate, median ACV, and median sales cycle.
- Check whether wins cluster in Tier 1 and losses in Tier 3. If not, adjust thresholds and re-run.
- Iterate until the model shows a clear correlation between tier and deal outcome.
05
Push tiers into CRM and route GTM effort
Write the tier score back to every CRM account record and use it to route acquisition budget, SDR focus, and campaign targeting.
- Use n8n or a native Clay integration to write the tier field into Salesforce or HubSpot.
- Make the tier visible on account, contact, and opportunity layouts so reps see it in-context.
- Route paid ads and ABM campaigns exclusively to Tier 1 accounts.
- Assign SDRs to manually prospect Tier 1, and send Tier 2 and 3 into automated sequences.
- Align messaging, case studies, and positioning content around the Tier 1 profile.