Which channel, message, or AI-generated asset deserves credit when a customer sees five touchpoints before buying? In this module, you will learn to assign conversion credit across AI-assisted customer journeys using several attribution models and to choose the one that fits your business. By the end, you will produce an attribution comparison report that shows how first-touch, last-touch, linear, time-decay, position-based, and Markov chain models redistribute credit among AI-generated and human-created touchpoints, along with a justified model recommendation. You will use path analysis, removal effects, Shapley value concepts, lookback window settings, and spreadsheet templates, and you will consider the limits created by cookie loss, walled gardens, and consent rules. You will also compare results with GA4 data-driven attribution and use an AI assistant to explain why credit shifts between models. Examples include AI-written nurture emails, generated social ads, and chatbot conversations that appear mid-journey. This module is designed for marketing managers, performance marketers, channel owners, and marketing analysts who report on channel and content performance. You need basic spreadsheet skills and familiarity with digital marketing channels; no coding or advanced statistics is required. You will take away a completed attribution comparison report, a model selection checklist, and the confidence to explain and defend your crediting approach in a budget conversation with channel owners and leadership.

Would those sales have happened anyway, even without your AI-generated campaign? In this module, you will learn to answer that question with uplift testing, which measures the incremental impact that attribution models cannot prove. By the end, you will design a geo-lift or holdout study for an AI campaign and calculate incremental conversions, incremental return on ad spend, and cost per incremental acquisition from the results. You will use audience holdouts, platform conversion lift studies, matched-market geo experiments, synthetic control, and difference-in-differences, along with the open-source GeoLift package and a spreadsheet incrementality calculator. You will also use an AI assistant to stress-test your design for contaminated controls and seasonality. Examples include always-on AI-generated social ads, AI-personalized email programs, and regional rollouts of generated video creative where user-level randomization is not possible. This module is designed for performance marketers, marketing analysts, media buyers, and marketing managers who need to justify spend with causal evidence. You need basic familiarity with campaign metrics such as conversions and ROAS; no coding is required, and you do not need any other module first. You will take away a completed incrementality study design, a lift readout with incremental ROAS and cost per incremental acquisition, and a method selection guide that helps you choose the right uplift test for every channel and budget level.

How do you find the customers who will buy only because they saw your AI-personalized offer, rather than those who would buy anyway? In this module, you will learn to build uplift models that predict the incremental effect of a marketing treatment for each customer. By the end, you will build and evaluate an uplift model on campaign data and produce a target list ranked by predicted incremental impact, with a cutoff chosen from a Qini curve. You will use the four uplift segments of persuadables, sure things, lost causes, and sleeping dogs, the two-model and class transformation approaches, Qini and uplift curves, and prepared Python notebooks that run in Google Colab with the open-source scikit-uplift library, plus an AI assistant that explains the code and helps you read the curves. Examples include AI-personalized discount emails, generated retention offers, and dynamic website messages where untargeted outreach wastes budget or annoys customers. This module is designed for marketing analysts, CRM and lifecycle marketers, and growth managers who decide who should receive a campaign. You need basic spreadsheet skills and an understanding of treatment and control groups; you will run prepared code rather than write it. You will take away a scored customer target list, an uplift model evaluation summary, and targeting rules that reduce wasted incentives and protect customer experience.

How much revenue do your AI-generated content and media investments really drive when user-level tracking is incomplete? In this module, you will learn to build a marketing mix model that estimates the contribution and return of each channel, including AI content production, from aggregate weekly data. By the end, you will fit a marketing mix model with adstock and saturation effects, estimate ROI and marginal ROI for each channel, and recommend a budget reallocation based on response curves. You will use regression-based marketing mix modeling, adstock carryover, diminishing returns curves, Bayesian priors, and calibration with lift test results, working in prepared notebooks built on Google Meridian or Meta Robyn alongside a spreadsheet planning template. An AI assistant will help you propose priors and translate response curves into a budget memo. Examples include paid social with AI-generated creative, AI-assisted SEO content output, email volume, and seasonal promotions. This module is designed for marketing analysts, media planners, and marketing managers who allocate budgets across channels. You need comfort with spreadsheets and basic familiarity with regression concepts such as trend and coefficients; you will run prepared code rather than write it. You will take away a fitted model summary, channel response curves, and a data-backed budget reallocation recommendation for AI content and media spend.

What do you report when attribution, lift tests, and your marketing mix model each tell a different story about AI content? In this module, you will learn to combine these methods into one consistent view of the business value that generative AI creates. By the end, you will build an AI content ROI scorecard that reconciles attribution, incrementality, and marketing mix estimates using calibration factors and reports incremental revenue, cost per incremental outcome, and a confidence rating for each channel. You will use the unified marketing measurement approach, incrementality calibration multipliers, a measurement confidence ladder, scenario-based KPI forecasting, and a spreadsheet scorecard template, with an AI assistant to draft scenarios and an executive summary that you verify against the numbers. Examples include reconciling platform-reported conversions with lift results, adjusting attribution credit for AI-generated ads, and forecasting revenue when AI content volume doubles. This module is designed for marketing leads, marketing analysts, growth managers, and finance partners who must agree on one number for AI content ROI. You need basic spreadsheet skills and familiarity with common marketing metrics such as ROAS and conversion rate; no coding is required. You will take away a triangulated ROI scorecard, a calibration worksheet, and a KPI forecast that connect generative AI outputs directly to revenue, profit, and growth targets.