Marty Smithson

Measurement

Working demos, not claims

Five small instruments that show how I think about marketing science: mix models, geo tests, experiment sizing, attribution, and the executive read. Every figure inside them is synthetic and generated in your browser. None of it is client or employer data.

martysmithson04@gmail.comLinkedInResume PDFMcfly Analytics Shopify App

  1. Demo · synthetic dataMMM budget reallocationWhere should the next dollar go?Channel response curves with saturation. Move spend between channels and watch predicted revenue, blended MER, and each channel's marginal return update live.
  2. Demo · synthetic dataGeo test readoutDid the spend cause the sales?Test markets against a pre-period-scaled control. A lift estimate with an uncertainty band and a keep, cut, or keep-testing call against break-even.
  3. Demo · synthetic dataExperiment power calculatorHow big, and how long?Sample size per arm and test duration from the baseline, the minimum detectable effect, the variance, and your traffic.
  4. Demo · synthetic dataAttribution comparisonWho gets credit, and does it add up?The same synthetic customers credited three ways: last-click, data-driven (Shapley), and an MMM-style incremental view. Plus a reconciliation of paid, affiliate, and creator claims against real orders.
  5. Demo · synthetic dataExecutive MER dashboardWhat did we book against what we spent?Cash MER against platform ROAS over time, with a break-even line set by contribution margin and a fail-closed week when the data isn't complete.

Method write-ups

  • From MTA to macro-measurementWhy user-level click paths stopped being a reliable budget source, dated and sourced, and what replaced them: modeled conversions, MMM, lift tests, and MER.
  • MMM workflowStart from the budget decision, build the data honestly, let experiments anchor the model, and present a recommendation with its range.
  • Designing a geo testPre-register the decision, pick markets that track each other, size the test from real pre-period noise, and read the interval against break-even.
  • Reconciling platform vs cashDefine one source of truth for sales and spend, explain the gap line by line, de-duplicate claims, and report cash MER with platform ROAS as context.