1300+ MQLs in 15 Days Using Lead Scoring

The Problem

Launching a new medical device product into an existing database is a signal problem, not a targeting one. Thousands of contacts existed across specialties and funnel stages, but nothing identified who was genuinely interested in this specific product. Without that signal, broad outreach was the only option — driving unsubscribe risk, inflated CPL, and low-quality leads during the most critical 15 days of a launch.

The Goal

Generate a high volume of product-specific MQLs within the first two weeks of launch — without burning the database — and build a repeatable scoring framework deployable across future product launches.

Strategy

Built a two-layer scoring model in Adobe Marketo: a Fit Score (job title, specialty, account type, geography) and an Engagement Score (page visits, content downloads, email interaction, webinar attendance). Combined into a composite score defining four Product Interest Groups (PIGs): High Interest, Moderate Interest, Low Interest, and Unscored — each with distinct messaging, cadence, and sales handoff thresholds. Critically, the model was configured before launch so pre-launch awareness activity passively populated scores, making the High Interest segment actionable on day one.

Execution

Configured Smart Lists, Score Fields, and triggered campaigns to automate scoring. Behavioral score decay was applied to contacts inactive for 90+ days. On launch day, the model ran against the full database: the High Interest PIG received the first outbound sequence within 24 hours. The Moderate Interest group entered a product education track before sales handoff. The full framework was then templatized — with configurable weights and thresholds — for rapid deployment across future launches.

Results

1,300+ MQLs generated in 15 days — a 3× increase in MQL velocity versus prior launches that used broad outreach. SAL rate was significantly higher because contacts were pre-qualified by fit and behavior before the first sales touch. The PIG framework gave sales a tiered, intent-ranked view of the database. The model was subsequently scaled across five additional product lines, consistently compressing time-to-first-MQL from launch day one.

What I'd Do Next

Move from rules-based to predictive scoring using historical conversion data, so the model learns which signals actually predict MQL-to-opportunity conversion. Layer in third-party intent data (Bombora, TechTarget) to capture out-of-platform research activity and give the model visibility into prospect intent before they engage with owned channels.