P2-18: Unified Contact Scoring

Automatically score contacts based on engagement, behavior, and profile data.

---

Scoring Card

Dimension	Score	Rationale
Pain	3 / 5	No way to prioritize contacts. Sales can’t find hot leads.
Revenue	3 / 5	Improves targeting efficiency across all modules
Build	4 / 5	Scoring rules engine + time decay + segment integration
Moat	3 / 5	Scores become more valuable as more events feed in
Total	13 / 20

---

Classification

Platform

Platform Primitive

Contact Scoring is infrastructure that enriches every contact with a numerical health/engagement signal. Segments, sequences, nudges, and sales workflows all benefit from knowing “how engaged is this contact?”

---

The Pain It Kills

Growth and sales teams need to prioritize, but every contact looks the same in a flat list:

1. No prioritization — sales doesn’t know which trial users are most likely to convert. They call everyone or nobody.
2. Manual lead scoring — growth teams build SQL queries (“users who logged in 5+ times AND invited a teammate AND viewed pricing”) to identify hot leads. These queries break and are never maintained.
3. Enterprise tools are overkill — HubSpot lead scoring requires Marketing Hub Professional ($800/mo). Salesforce Einstein requires Enterprise edition ($150/user/mo). Both are designed for enterprise sales processes, not PLG.
4. No decay — a user who was active 6 months ago but hasn’t logged in since still shows as “engaged” because there’s no time decay.
5. At-risk users are invisible — growth teams can’t identify users whose engagement is declining before they churn.

Real scenarios:

• A SaaS product has 500 trial users. The growth team wants to focus their limited time on the 50 most likely to convert. Today: they export login frequency from analytics, merge with feature usage from the database, and manually rank in a spreadsheet. Takes 2 hours and is outdated by the time it’s done.
• A customer success manager wants to know which paying customers are at risk of churning. Today: gut feel based on last support ticket.
• A product team wants to show upgrade nudges only to users with high engagement scores. Today: impossible without custom code.

---

What It Does

Contact Scoring automatically calculates and maintains a numerical score for every contact based on configurable rules:

• Event-based scoring — assign points for actions: login (+2), invite teammate (+10), create project (+5), view pricing (+15), complete onboarding (+20).
• Property-based scoring — bonus points for profile attributes: paid plan (+30), company_size > 50 (+10), role = decision_maker (+20).
• Time decay — scores decay over time. A login 30 days ago is worth less than a login today. Configurable decay rate.
• Score history — full history of score changes per contact, enabling trend analysis (“engagement increasing” vs “engagement declining”).
• Segment integration — create segments based on score thresholds (“hot leads: score > 80,” “at-risk: score declining > 20 points in 7 days”).

---

Competition & What We Replace

Tool	Price	Limitation
HubSpot lead scoring	$800+/mo (Professional)	Enterprise pricing. Sales-focused, not PLG.
Salesforce Einstein	$150+/user/mo	Enterprise. AI-powered but requires massive data.
Custom SQL queries	Engineering time	Fragile, no decay, no real-time updates.
Mixpanel engagement score	Analytics plan	Analytics-only. Can’t trigger actions.
GrowthOS Contact Scoring	Included	Event + property + decay, real-time, powers segments and sequences

---

Moat & Defensibility

The moat is data breadth:

• Contact Scoring consumes events from every GrowthOS module: email opens, nudge clicks, referrals, survey responses, onboarding progress, feature usage.
• The more modules a tenant activates, the more accurate the scoring becomes.
• Standalone scoring tools only have access to the data you send them. GrowthOS scoring has access to everything.
• Score data feeds into segments, which power sequences, nudges, and review prompts — creating a virtuous cycle that makes the scoring more useful over time.

---

Interoperability Advantage

Contact Scoring consumes data from every module and feeds scores into Segments, which power every outreach module.

---

What Ships

1. Scoring rule builder — dashboard UI to configure event weights and property bonuses
2. Event-based scoring — assign points for any GrowthOS event (login, feature use, email open, etc.)
3. Property-based scoring — bonus points based on contact properties (plan, role, company size)
4. Time decay — configurable decay rate so scores reflect recent engagement, not historical
5. Score history — full timeline of score changes per contact with trend indicators
6. Segment integration — use score thresholds in Segment Builder rules

---

What Does NOT Ship

• Predictive scoring — ML-based “likelihood to convert/churn” scoring is planned for P4.
• Custom ML models — no ability to train custom scoring models on tenant data.
• Intent signals from external data — no ingesting third-party intent data (Bombora, G2 intent, etc.).
• Multi-score models — one score per contact in P2. Multiple scores (engagement, fit, intent) are a future enhancement.

---

Build vs Buy

Build (chosen)

• Scoring rules engine is similar to Segment Builder rules (shared infrastructure)
• Deep integration with Event Bus is essential — no off-the-shelf tool plugs in
• Time decay requires a background job but is straightforward
• Estimated: 2 weeks

Buy

• MadKudu ($500+/mo) provides scoring but requires data export/import pipeline
• HubSpot scoring requires adopting the full HubSpot stack
• No standalone scoring tool integrates with GrowthOS events natively

---

Dependencies

Dependency	Phase	Status	Notes
Contact Graph	P1	Required	Contact properties for property-based scoring
Event Bus	P1	Required	All module events for event-based scoring
Segment Builder	P2	Optional	Score-based segments (bidirectional integration)