Resources
Explore the book, whitepapers and in-depth notes on the methods that inform Kuwalyst.
The book
BookMarketing Science in Python
A practical guide to causal inference, media mix modelling, CLV, pricing and forecasting in Python, available as a personalised PDF.
Get the free PDF →Whitepapers
Media planning with AI
How agentic systems build, monitor and tune media plans, and the guardrails that keep every number tied to evidence.
Coming soonMethod explainers
Bayesian media mix modelling
A media mix model estimates how much each channel contributes to sales while accounting for carryover and saturation. A Bayesian approach makes prior assumptions explicit and returns a range of plausible outcomes rather than a single point estimate.
8 min readMethodExcess share of voice
Excess share of voice compares a brand's share of category advertising with its share of category sales. A positive gap is generally associated with market-share growth, while a negative gap is associated with decline, making it a useful guide to brand-budget size.
6 min readMethodGeo-lift experiments
A geo-lift experiment compares regions where advertising is paused with similar regions where it continues. It measures incremental effect where user-level tracking is not possible and can calibrate a media mix model against observed results.
8 min readMethodGeometric adstock
Advertising can continue to have an effect after it has run. Geometric adstock models how much of this week's impact remains next week as a fixed proportion that decays over time.
6 min readMethodHill saturation
Additional investment often delivers less incremental effect than the initial spend. The Hill function models this relationship with a half-saturation point and a curve shape that describes how quickly returns increase and then level off.
6 min readMethodMarginal incremental ROAS
Marginal incremental ROAS focuses on the return from the next euro of spend. It considers only sales that would not otherwise have happened, rather than the average return reported across all spend, so channels can be compared more meaningfully.
7 min readMethodResponse curves
A response curve estimates what a channel returns at different levels of spend, rather than only at its current level. It helps planners understand the likely return from the next budget decision instead of relying on an average return.
6 min readMethodThe audited benchmark registry
When a brand has little history, a plan may need to start with published benchmarks. Kuwalyst's registry traces each claim to its primary publisher, audits it and assigns a status that determines whether it can be used.
7 min read