fsbtactic-code/marketing-brain-skill / skills/marketing-brain
marketing-brain
FlaggedCreate, improve, research, or audit marketing strategy and execution using a bundled marketing library, verified live primary sources for volatile facts, and clearly labeled agent judgment. Use for full marketing systems, audits, landing pages and CRO, competitor intelligence, brand voice, audience and JTBD research, email sequences, SMM and social content systems, paid ads, headlines, copywriting, positioning, offers, launches, GTM, growth, funnels, retention, pricing, SEO, analytics, attribution, experiments, and consumer or buyer psychology. Trigger for English and Russian requests, including requests for ready-to-use assets.
assets/wiki/domains/measurement-analytics.md
View on GitHub---
page_id: "page_domain_measurement_analytics"
title: "measurement-analytics"
page_type: "domain"
updated_at: "2026-07-22"
status: "verified"
card_refs:
- "card_ad_effects_heterogeneous_by_firm"
- "card_ad_roi_measurement_noise"
- "card_google_ads_experiment_hygiene"
- "card_incrementality_control_group"
- "card_incrementality_precision_stratification"
- "card_short_video_ad_spillover"
source_refs:
- "src_berman_feit_incrementality_2024"
- "src_bytedance_ad_spillover_2025"
- "src_dai_kim_luca_digital_ads_2023"
- "src_google_ads_conversion_lift_2026"
- "src_google_ads_experiment_practice_2026"
- "src_lewis_rao_ad_returns_2015"
---
# measurement-analytics
## A randomized field experiment across restaurants found digital advertising effects varied by firm characteristics, with larger gains for some independent, higher-rated, and higher-organic-traffic businesses.
Segment ad tests by preexisting firm attributes and organic demand, and avoid applying an average result uniformly across accounts.
Card: [card_ad_effects_heterogeneous_by_firm](../cards/card_ad_effects_heterogeneous_by_firm.md)
Evidence: B, confidence 0.84
- [Which Firms Gain from Digital Advertising? Evidence from a Field Experiment](https://pubsonline.informs.org/doi/abs/10.1287/mksc.2023.1436), locator: Abstract, DOI 10.1287/mksc.2023.1436
## Across 25 large field experiments, Lewis and Rao found advertising ROI estimates were often extremely imprecise and argued that randomized trials add unbiased information while observational methods face serious selection bias.
Plan ad incrementality tests around detectable effect size, variance, sample size, and cost. Treat observational attribution as vulnerable to targeting selection bias.
Card: [card_ad_roi_measurement_noise](../cards/card_ad_roi_measurement_noise.md)
Evidence: A, confidence 0.92
- [The Unfavorable Economics of Measuring the Returns to Advertising](https://academic.oup.com/qje/article-abstract/130/4/1941/1914592), locator: Abstract, DOI 10.1093/qje/qjv023
## Google Ads recommends a clear hypothesis, testing one variable at a time, selecting success metrics before the test, limiting base-campaign changes, and keeping experiment records.
Before launching an ad experiment, state one business-linked hypothesis, isolate the changed variable, choose one or two decision metrics, avoid contaminating the base, and log the result.
Card: [card_google_ads_experiment_hygiene](../cards/card_google_ads_experiment_hygiene.md)
Evidence: A, confidence 0.97
- [Test with confidence with the Experiments page](https://support.google.com/google-ads/answer/7281575?hl=en), locator: Sections 1 through 3: Set a clear hypothesis; Create your experiment; Analyze results
## Google defines Conversion Lift as a controlled experiment that separates audiences into treatment and control groups to estimate conversions caused by ad exposure.
Use a treatment and unexposed control design when the decision requires causal incremental conversions rather than attributed conversions alone.
Card: [card_incrementality_control_group](../cards/card_incrementality_control_group.md)
Evidence: A, confidence 0.98
- [About Conversion Lift](https://support.google.com/google-ads/answer/12003020?hl=en), locator: Sections: About Conversion Lift; How Conversion Lift works
## Berman and Feit report that latent stratification reduced average-treatment-effect variance by 30% to 60% in five catalog experiments under favorable identification conditions.
For noisy lift tests, consider preplanned heterogeneity models only when the strata model is identifiable, documented, and validated against a standard estimator.
Card: [card_incrementality_precision_stratification](../cards/card_incrementality_precision_stratification.md)
Evidence: B, confidence 0.83
- [Latent Stratification for Incrementality Experiments](https://pubsonline.informs.org/doi/abs/10.1287/mksc.2022.0297), locator: Abstract, DOI 10.1287/mksc.2022.0297
## A 2025 randomized ByteDance field experiment found substantial off-platform conversion spillover, which materially changed campaign cost-per-conversion estimates in the studied automobile campaign.
When evaluating short-form video ads, test off-platform outcomes where possible and do not assume platform-reported conversions capture the full effect.
Card: [card_short_video_ad_spillover](../cards/card_short_video_ad_spillover.md)
Evidence: B, confidence 0.82
- [Is the Money Spent on Short-Form Video Social Platforms Worth It?](https://pubsonline.informs.org/doi/10.1287/mksc.2023.0575), locator: Abstract, DOI 10.1287/mksc.2023.0575
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |