Lochanreddy MallakuntaAnalytics portfolio
MarketingIndependentIndependent portfolio project

Marketing Attribution and Customer Segmentation

An analysis framework that distinguishes channel quality and retained customer value from low-quality inexpensive conversions using synthetic campaign data.

Tools: SQL · Python · Excel · Power BI

Overview

Independent marketing analytics project focused on channel comparison, segmentation, retention concepts, and clear separation between rule-based attribution, experiments, and correlation.

Business problem

Which channels and customer segments create valuable, retained customers rather than inexpensive but low-quality conversions?

Analysis and approach

  • Define acquisition and retention KPIs
  • Compare channels on value, not only CPA
  • Segment customers by behavior or value
  • Label attribution evidence strength clearly

Skills: Campaign performance · Segmentation · Cohort analysis · Attribution concepts

Evidence and visuals

  • Placeholder for marketing attribution evidence

    Evidence placeholder — synthetic channel and cohort visuals pending.

Findings and outcome

  • Low CPA does not guarantee valuable retained customers.
  • Rule-based attribution should not be described as causal lift.
  • Segment-level retention changes the budget conversation.

Outcome: An analysis framework that distinguishes channel quality and retained customer value from low-quality inexpensive conversions using synthetic campaign data.

Recommendations

  1. 1

    Report channel quality with retention or LTV proxies beside CPA.

  2. 2

    Label last-touch and other rule-based models as observational.

  3. 3

    Use experiments when incremental lift decisions are high stakes.

Project details

Dataset / source
Synthetic campaign and customer dataset
Data classification
Synthetic
Limitations
  • Synthetic data cannot prove real campaign ROI.
  • Observational attribution is not causal evidence.