Lifecycle & Scoring

Multi-Dimensional Scoring & Lifecycle Redesign

A single, global scoring model produced inaccurate prioritization across multiple products. I built portfolio‑specific scoring models (behavioral, demographic, and firmographic), added score decay and eligibility checks and re‑architected HubSpot lifecycle workflows into modular components. The result was a 20 % increase in MQL accuracy and a 30 % rise in MQL‑to‑SQL conversion.

The Problem

At a high-growth SaaS business, the marketing team depended on a single global scoring model to qualify leads for multiple product lines. While easy to manage, the model produced inaccurate scores, misaligned prioritization, and inconsistent funnel progression across solutions. Different products targeted different ICPs, buying motions, and engagement patterns, but the scoring logic applied the same weights and criteria to all contacts, creating mismatches between score and actual intent.

Behavioral scoring was limited to a “one-time credit” mechanic: if a user clicked an email, downloaded content, or visited a page, they received a single point assignment regardless of ongoing behavior. This prevented the score from reflecting recency, frequency, or cumulative intent. Demographic and firmographic signals were underutilized, and the lack of score decay inflated scores over time, making it impossible to distinguish fresh intent from legacy interactions.

Lifecycle processing in HubSpot also reflected the constraints of the old model. A single, deeply nested workflow attempted to manage all lifecycle transitions. This design increased error rates, restricted iteration, and created cascading downstream issues when exceptions or new motions were added.

The mismatch between business complexity and system architecture resulted in inconsistent MQL quality, inflated lead volume, limited funnel visibility, and a qualification process that no longer supported the organization’s growth.

The scoring model could not accurately qualify leads because it treated every portfolio, persona, and behavior the same.

the system was designed for simplicity, not for accuracy, intent, or ICP alignment.

The Solution

Multi-Dimensional Portfolio-Specific Scoring Models

I designed and implemented independent scoring models for each product portfolio, aligning behavioral, demographic, and firmographic signals to product-specific qualification patterns. This matrixed approach allowed each model to assess intent based on its unique buying motion without duplicating global logic. Portfolio-specific scoring enabled accurate prioritization across segments and reduced noise in the funnel.

Behavioral Scoring Redesign with Ongoing Activity Tracking

I replaced one-time engagement points with dynamic behavioral scoring that reflects true buying intent. New mechanics included: cumulative scoring for repeat behaviors (email clicks, page visits, content engagements), event-based scoring for high-value actions, weighting logic aligned to funnel stage influence. This eliminated static score inflation and ensured that score reflected real activity over time.

Score Decay, Ineligibility Processing & Quality Controls

To prevent inflated scores and ensure recency mattered, I introduced score decay logic that gradually reduced score for inactive contacts. I also built ineligible processing that automatically suppressed free emails, invalid engagement patterns, and non-qualified segments, improving MQL accuracy and decreasing unqualified volume before handoff.

Modular Lifecycle Redesign in HubSpot

I re-engineered the lifecycle architecture from one monolithic process into a modular ecosystem of independent workflows, each responsible for a single lifecycle stage. This design enabled precise control over progression rules, improved visibility for operations and sales, reduced regression errors and duplicate processing, made lifecycle adjustments low-risk and scalable.

Turned noisy scoring into

a modular, portfolio-aware qualification engine.