Lead Scoring & Routing — Fit × Engagement, Thresholds, SLAs¶

For factory operators. The funnel benchmarks in benchmarks-as-priors show MQL→SQL is the steepest leak (~39% on average). That leak is a qualification-and-routing problem, and this document is its fix: score leads on the two dimensions that predict conversion, route the hot ones fast, and enforce the marketing-sales contract that stops good leads from rotting.

The thesis: a lead score is a prioritization tool, not a truth. Its only job is to get the right leads to the right rep fast, with a feedback loop that keeps the model honest.

1. The Scoring Model — Fit × Engagement¶

Every lead is scored on two independent dimensions:
- Fit (demographic / firmographic): static profile match to the ideal-customer-profile — company size, revenue, industry, job title, seniority, geography, tech stack. Does this lead match who we sell to?
- Engagement (behavioral): activity intensity — and high-intent actions are weighted heavily (a single demo request outweighs a dozen email opens). Is this lead actively buying?

The two-by-two is the point: high-fit + high-engagement is the only quadrant worth a fast sales touch. High-engagement + low-fit wastes rep time; high-fit + low-engagement needs nurture, not a call.

Model types: demographic (fit), behavioral (engagement), predictive (ML on historical CRM conversions to learn which attributes correlate with closed-won), and hybrid — most teams land on hybrid, balancing automation with human judgment.

2. Thresholds — Defining MQL and SQL¶

A score becomes an action at a threshold. (Reported, secondary: common practice sets the MQL threshold around 60–80 points — roughly the top 15–20% of the database — with a separate, higher SQL-handoff threshold and a "hot" band around 71+.) The thresholds are not sacred numbers; they're calibrated so the MQL rate matches sales capacity and the expected conversion-by-score-band holds up. Recalibrate from actual conversion data — this is where marketing-attribution and the event-tracking-schema feed back.

3. Routing & the Marketing-Sales SLA¶

A score is worthless if the lead sits. Routing turns thresholds into assignment:
- Threshold = routing trigger: crossing "hot" (e.g. 71+) auto-assigns by territory, rep availability, and capacity.
- Speed SLA: an MQL should reach a specific rep within ~5 minutes — the same speed-to-lead finding that recurs across the KB (contacting within an hour ≈ 7× the conversion of waiting a day).
- Acceptance SLA + feedback loop: the rep has ~48 hours to accept or reject, with mandatory feedback on rejections. That feedback is the training signal that keeps scoring honest and stops the classic marketing-sales blame cycle.
- Shared definitions: marketing and sales agree on MQL/SQL definitions, expected conversion by score band, and the SLAs governing each — one contract, jointly owned.

4. Factory Integration¶

Lead scoring and routing is the funnel's connective tissue and almost entirely agent-ownable: computing fit × engagement scores from event-tracking-schema and identity-resolution data, firing routing on threshold crossings, and enforcing SLA timers and feedback capture are all mechanical. The human owns the accept/reject judgment and the shared-definition negotiation. It directly attacks the MQL→SQL leak the benchmarks-as-priors data exposes, and the rejection-feedback loop is a small, continuous experiment-loop that recalibrates thresholds against real conversion.

Provenance Note¶

Synthesized from 2026 RevOps practitioner sources (secondary, NOT single-primary fetched): the fit × engagement model, the four model types, and all threshold/SLA figures (MQL ~60–80 / top 15–20%, hot ~71+, ~5-minute routing, ~48-hour acceptance). Figures are reported common practice, not independently verified or universal.
First-principles / KB-internal: tying scoring to the funnel-leak benchmark, the data layer, and the feedback loop as a mini experiment loop.

Sources¶

NC Squared / Pedowitz Group / Scalarly — Lead scoring models & B2B templates (secondary)
Prospeo — Marketing Qualified Lead Criteria: Scoring & SLA (2026) (secondary)

Concepts¶

Extracted from this source: lead-scoring-model · lead-routing-sla