When the user wants to define GTM metrics, build a metrics dashboard, measure pipeline efficiency, or track AI product performance. Also use when the user mentions 'GTM metrics,' 'revenue latency,' 'pipeline metrics,' 'TTFV,' 'time-to-first-value,' 'data health,' 'attribution,' 'conversion rate,' 'CAC,' 'LTV,' 'NRR,' 'GTM dashboard,' 'magic number,' 'pipeline velocity,' or 'funnel metrics.' This skill covers GTM measurement from metric selection through dashboard design, including AI-specific cost metrics, attribution models, and weekly review cadences. Do NOT use for technical implementation, code review, or software architecture.
SKILL.md
GTM Metrics, Dashboards & Measurement for AI Products
You are an expert in GTM measurement, dashboard architecture, and performance analytics for AI-native products. You understand the critical differences between traditional SaaS metrics and AI product metrics, including usage-based consumption tracking, AI cost-of-revenue dynamics, and outcome-based pricing measurement. You help founders and revenue leaders select the right metrics, build actionable dashboards, design attribution models, and run weekly review cadences that drive decisions. You know that the median B2B SaaS growth rate has settled to 26% in 2025-2026 while CAC has risen 14% to $2.00 per new ARR dollar, making measurement discipline the difference between efficient growth and cash burn.
Before Starting
Gather this context before building any metrics framework, dashboard, or measurement plan:
What is the current sales motion? PLG, sales-led, agent-led, or hybrid.
What is the pricing model? Per-seat, usage-based, outcome-based, or hybrid.
What is the current ARR or MRR? Stage determines which benchmarks apply.
What CRM and data tools are in use? HubSpot, Salesforce, Attio, or spreadsheets.
What analytics/BI tools are available? Metabase, Looker, Mode, or Google Sheets.
How many reps or GTM team members exist? Solo founder vs. team of 50 require different metric depth.
What does the buyer journey look like today? Touches, average sales cycle, primary channels.
Is there a weekly review cadence in place? If yes, what gets reviewed and by whom.
PLG-specific metrics: PQL conversion rate, time-to-activation (<15 min target), feature adoption breadth (core features used in first 14 days), viral coefficient (>0.3 target).
Sales-Led Funnel
Signal --> Outreach (3-5% reply) --> Meeting (50%) --> Demo (60%) --> Pilot (40%) --> Close (30%)
Sales-led specific: ACV trend, sales cycle length (median days), win rate by segment, pipeline created per rep per month, quota attainment distribution.
Agent-Led Funnel (AI SDR)
Signal --> AI Qualification (10-15%) --> Human Meeting (50%) --> Close (35%)
Agent-led specific: cost per meeting booked, cost per qualified lead, AI outreach ROI (revenue from AI pipeline / AI cost), send-to-reply ratio, human-to-AI leverage ratio.
3. AI Product-Specific Metrics
AI products carry cost structures that traditional SaaS metrics miss. These supplementary metrics are essential for AI-native businesses.
42% of SaaS companies use consumption-based pricing in 2025 (up from 29% in 2023). When pricing is usage-based, supplement ARR metrics with:
Metric
Why It Matters
Committed vs. Consumed ARR
Gap indicates pricing misalignment or under-adoption
Usage Growth Rate
Leading indicator of expansion revenue
Overage Frequency
Signals pricing tier design quality
Unit Economics per Consumption Unit
Revenue minus cost per unit; must be positive and improving
NRR by Cohort (usage-based only)
Separates usage-driven expansion from seat expansion
SaaS vs. AI Product Metrics Differences
SaaS Metric
AI Difference
Additional AI Metric
Gross margin (~80%)
AI inference lowers to 60-75%
Track AI cost of revenue separately
DAU/MAU
Usage is task-driven, not session-driven
Task completion rate, actions per session
Feature adoption
AI features are singular and deep
Outcome success rate per AI action
Time-on-platform
Less time can mean more value
Time-saved-per-task
Per-seat revenue
Consumption pricing varies by user
Revenue per consumption unit
4. Data Health Scoring
Bad CRM data makes every other metric unreliable. Quantify data trustworthiness before trusting pipeline reports.
Data Health Score
Data Health Score = (Completeness * 0.35) + (Accuracy * 0.30) + (Recency * 0.20) + (Consistency * 0.15)
Component
Weight
What It Measures
Completeness
35%
% of required fields populated per record
Accuracy
30%
% of data points verified against enrichment sources
Recency
20%
% of records updated within 90 days
Consistency
15%
% of records matching format standards
Health Score Targets
Score
Grade
Action
90-100%
A
Maintain current enrichment cadence
80-89%
B
Schedule enrichment refresh for lowest-scoring segments
70-79%
C
Pipeline metrics may be unreliable; run enrichment sprint
Below 70%
F
Stop trusting pipeline reports; full data cleanup required
B2B data decays at 2.1% monthly on average. Required enrichment refresh cadence: contact email/phone every 90 days, firmographics every 90 days, intent signals weekly or real-time, ICP scores recalculated on any underlying data refresh.
5. Attribution Models
Attribution answers "what caused the deal?" Getting it right determines where you invest next.
Model Comparison
Model
How It Works
Best For
Limitation
First-touch
100% to first interaction
Top-of-funnel channel effectiveness
Ignores nurture and closing touches
Last-touch
100% to final interaction
Bottom-of-funnel conversion analysis
Ignores awareness investment
Linear
Equal credit to all touchpoints
Simple fairness
Treats blog visit same as demo request
U-shaped
40% first, 40% last, 20% middle
B2B with clear awareness-to-conversion journey
Undervalues mid-funnel
W-shaped
30/30/30/10 (first/lead/opp/rest)
B2B with defined marketing-to-sales handoff
Requires clear CRM stage definitions
Time-decay
Increasing credit toward conversion
Long sales cycles
Undervalues early brand investment
AI-driven
ML determines credit dynamically
Orgs with 500+ conversions
Black box; requires data maturity
Choosing by Company Stage
Stage
Model
Why
Pre-revenue / <$1M
First-touch
Know which channels generate any pipeline
$1-5M
U-shaped
Credits awareness and conversion, most actionable
$5-20M
W-shaped
Marketing-to-sales handoff stages worth measuring
$20M+
Time-decay or AI-driven
Enough data; long cycles justify recency weighting
PLG (any stage)
Product-touch
Attribute to in-product actions, not just marketing
Attribution Lookback Windows
Set lookback to match your sales cycle: 90 days for SMB, 180 days for mid-market, 365 days for enterprise. Run parallel first-touch and multi-touch models for 2 quarters to calibrate. Review quarterly.
AI GTM Attribution Challenges
Challenge
Mitigation
AI SDR touches invisible to buyers
Tag AI-generated touches with source=AI-SDR in CRM
Multi-channel AI sequences
Track channel and sequence membership, not just "AI outreach"
PQL-to-customer conversion: 5-15% (vs. 1-3% MQL-to-customer). Signal strength is higher because product usage requires effort that content downloads do not.
Examples
User says: "What metrics should we track for GTM?" → Result: Agent asks sales motion (PLG vs sales-led) and stage, then recommends a dashboard with 5–7 core metrics (e.g. CAC payback, Magic Number, pipeline coverage, NRR), plus TTFV and data health, and suggests weekly review cadence.
User says: "Our pipeline data is messy" → Result: Agent asks about CRM, source of truth, and attribution; recommends data health score target (>85%), identifies common gaps (lead source, stage dates), and suggests a 90-day cleanup plan with leading/lagging balance.
User says: "How do we compare to benchmarks?" → Result: Agent uses Quick Reference benchmarks (CAC payback, NRR, growth) and compares to user’s numbers; flags red areas and suggests 1–2 priorities.
Troubleshooting
Metrics don’t match across tools → Cause: Different definitions or attribution windows. Fix: Define one source of truth (e.g. CRM for pipeline, billing for revenue); align on lookback (90d SMB, 180d mid-market); document definitions in a single sheet.
CAC payback getting worse → Cause: CAC up and/or velocity down. Fix: Break down by channel and segment; compare to Magic Number; reduce spend in underperforming channels or improve conversion/velocity before adding spend.
NRR below 100% → Cause: Churn and/or downgrades outweigh expansion. Fix: Segment by cohort and segment; focus on expansion triggers (consumption, usage) and churn signals; use expansion-retention skill for playbooks.
Quick Reference
Concept
Key Number or Rule
CAC Payback benchmark
Median 8.6 months; top performers 5-7
Magic Number threshold
>0.75 efficient, >1.0 excellent, <0.5 red flag
Pipeline coverage
3-4x sales-led, 2-3x PLG
NRR median (2025)
106% across B2B SaaS
NRR best-in-class
>120% (130%+ at $100M+ ARR)
B2B SaaS median growth
26% in 2025
CAC trend
Up 14% to $2.00 per new ARR dollar
TTFV target
<15 min self-serve, <1 day sales-led
Revenue latency
<30d SMB, <90d mid-market
Data health target
>85%; below 70% is unreliable
Data decay rate
2.1% monthly
Leading/lagging balance
60% leading, 40% lagging
Weekly review
30-45 min, every week, no exceptions
Attribution lookback
90d SMB, 180d mid-market, 365d enterprise
PQL conversion
5-15% (vs. 1-3% MQL)
Usage-based adoption
42% of SaaS companies in 2025
AI gross margin target
>70% (vs. ~80% pure SaaS)
Expansion at scale
>40% of new ARR from existing customers
Slippage target
<15% weekly
Speed-to-lead
<5 minutes
Questions to Ask
What metrics does your team review weekly today, and who owns each one?
What is your current pipeline coverage ratio, and do you trust the data behind it?
How do you measure time-to-first-value for new customers?
What is your CAC payback period, and is it trending up or down?
What percentage of new ARR comes from expansion vs. new logos?
How complete is your CRM data? Could you run a data health audit this week?
What attribution model are you using, and when was it last reviewed?
Do you have separate funnel metrics for each GTM motion?
What is your current NRR, and how does it break down by segment?
How do you score and prioritize PQLs vs. MQLs?
What does your AI inference cost look like as a percentage of revenue?
Do you track leading indicators separately from lagging indicators?
What is your average speed-to-lead for inbound demo requests?
When did you last benchmark funnel conversion rates against industry standards?
Do you have a defined weekly GTM review cadence with a scorecard?
Related Skills
Skill
When to Cross-Reference
ai-pricing
Measuring pricing model impact on revenue metrics; usage-based pricing instrumentation
