Marketing Analytics with Fexingo: Data, Attribution, and Measuring Campaign Performance

In this episode, Lucas and Luna dive into the hidden inefficiencies in digital advertising, using a real-world example from a mid-sized retailer that ran a three-week geo-based incrementality test. They reveal how the test showed that 40% of paid search spend was cannibalizing organic conversions, saving the company $200,000 annually. The discussion covers the mechanics of holdout groups, the pitf...

Lucas and Luna unpack a little-known measurement technique called ghost ads, which lets marketers isolate the true incremental impact of brand campaigns. They walk through how Procter & Gamble used ghost ads to uncover that nearly 30 percent of their digital brand spend was reaching people who would have bought anyway. The hosts explain the mechanics: serving a control group a public-service ad in...

Lucas and Luna dive into why incrementality testing is the only reliable way to measure true ad impact, using Amazon's 2021 shadow-ban experiment and Uber's 2024 geo-testing results as concrete examples. They explain how standard attribution models overcredit channels that merely appear coincident with conversions, while incrementality isolates the actual lift. They walk through the mechanics of a...

Episode 54 of Marketing Analytics with Fexingo. Lucas and Luna dive into why incrementality testing—running controlled experiments to measure true causal lift—is the gold standard for marketing ROI. They walk through a real case: how a DTC skincare brand used geo-based incrementality tests to discover that 40% of their Facebook ad conversions would have happened anyway. The hosts contrast this wit...

In Episode 53 of Marketing Analytics with Fexingo, Lucas and Luna delve into why standalone attribution models miss critical saturation effects, using a real-world case of a D2C brand that wasted $2 million on over-served channels. They explain how integrating media mix models (MMM) with attribution reveals diminishing returns and budget reallocation opportunities. Listeners learn about the concep...

Lucas and Luna unpack why even the best multi-touch attribution models can't answer the fundamental question: did the ad cause the sale? Using a real-world case study from a D2C mattress company, they walk through how incrementality testing — comparing a test group exposed to ads against a holdout group — revealed that 40% of attributed conversions would have happened anyway. They discuss the oper...

Episode 51 of Marketing Analytics with Fexingo tackles a blind spot in modern attribution: the failure to integrate media mix models with multi-touch attribution. Lucas and Luna explore a real case from a mid-size CPG brand that ran both models in parallel for six months and found a 23% discrepancy in channel ROI estimates. They unpack why siloed models lead to conflicting budget recommendations,...

Episode 50 of Marketing Analytics with Fexingo. Lucas and Luna explore a blind spot in most multi-touch attribution models: they treat each marketing channel as if it has unlimited capacity to drive conversions. Using the case of a DTC skincare brand that saw cost-per-acquisition triple after scaling Facebook spend past a threshold, they explain why saturation points exist, how to detect them usin...

Episode 49 of Marketing Analytics with Fexingo digs into a blind spot in most attribution models: paywall and registration-wall data. Lucas and Luna examine how the New York Times uses metered subscription analytics to measure content value far beyond clicks and scroll depth. They discuss why conversion-to-subscription metrics beat engagement proxies, how a paywall changes what you optimize for, a...

Episode 48 of Marketing Analytics with Fexingo. Lucas and Luna dig into a neglected layer of campaign measurement: the data quality score. Most marketers trust their attribution numbers without auditing the raw feed. But one CPG brand found that 23% of their ad interactions were misattributed because of a simple UTM parameter typo. The hosts walk through how to build a data quality scorecard — che...

In this episode, Lucas and Luna dive into the critical but often overlooked practice of data freshness monitoring in marketing attribution. They use the example of a mid-size e-commerce retailer whose attribution model broke because it was relying on stale conversion data. The hosts explain how latency in data pipelines—anywhere from 24 hours to a week—can lead to misallocated budgets and missed w...

Episode 46 of Marketing Analytics with Fexingo tackles a hidden killer of attribution accuracy: stale data. Lucas and Luna walk through a real case from a mid-market e-commerce brand whose last-click model showed email driving 40 percent of revenue — but the data feed was two weeks delayed on web sessions. They explain why data freshness matters more than model sophistication, how to set refresh S...

In this episode of Marketing Analytics with Fexingo, Lucas and Luna dive into a common pitfall in marketing mix modeling: failing to adjust for seasonality. Using the example of a mid-size e-commerce brand that saw a 20% sales lift in Q4 and wrongly attributed it to Facebook ads, they explain how naive models conflate seasonal demand with campaign effectiveness. Lucas walks through a concrete case...

Episode 44 of Marketing Analytics with Fexingo dives into a persistent flaw in how most brands measure campaign performance: they treat attribution models as static. Lucas and Luna explain why Bayesian updating — a technique borrowed from data science — can make attribution far more accurate by continuously adjusting credit as new data comes in. They walk through a concrete example: a $2 million r...

Lucas and Luna dig into a surprisingly common marketing analytics failure: attribution models that look sophisticated but are built on flawed data feeds. They use a real example from a mid-market DTC brand that spent six months optimizing toward a channel that, according to its own CRM, had never generated a single qualified lead. The episode walks through what a data feed quality audit actually l...

Episode 42 of Marketing Analytics with Fexingo. Lucas and Luna unpack the single biggest blind spot in modern marketing measurement: budget cannibalization from channels that look effective in your attribution model but are actually stealing conversions from other touchpoints. Using a concrete example from a fictional DTC brand called 'Bloom & Spruce', they walk through how a last-click model can...

In episode 41 of Marketing Analytics with Fexingo, Lucas and Luna unpack why traditional marketing mix models fall short when data is sparse or campaigns change fast. They walk through a concrete example: a mid-size consumer goods brand that wasted $2 million on TV ads because its frequentist model couldn't handle seasonality shifts. Lucas explains how a Bayesian framework — updating probabilities...

Episode 40 digs into a blind spot in multi-touch attribution models: how they systematically undervalue early-funnel content like blog posts and educational videos. Lucas and Luna examine a case from a B2B SaaS company that ran a six-month experiment comparing last-click vs. multi-touch attribution for its content program. The result? Multi-touch credited the final demo request to the email sequen...

Episode 39 of Marketing Analytics with Fexingo. Lucas and Luna unpack the growing role of data clean rooms in marketing attribution. Using the 2025 privacy changes to Apple's SKAdNetwork and Google's Privacy Sandbox as a backdrop, they walk through how a clean room lets advertisers join first-party data with a publisher's data without sharing raw user IDs. They cite a case where a major retailer r...

Episode 38 of Marketing Analytics with Fexingo. Lucas and Luna dig into geo-based incrementality testing — why running controlled experiments across geographic regions is the gold standard for measuring true ad lift. They break down a real example: how a national quick-service chain used city-level test-and-control markets to isolate the impact of a new digital campaign, uncovering that their mult...

Most marketers still rely on last-click attribution, but it systematically undervalues upper-funnel channels. In this episode, Lucas and Luna break down a real-world case from a $200 million DTC brand that switched from last-click to a Bayesian marketing mix model. They explain why the switch revealed that TV and podcasts were driving 40% of incremental revenue — not the 8% last-click showed — and...

Lucas and Luna dig into the viewability metric that the entire digital ad industry relies on — and why it's often misleading. Using a real example from a $2 million programmatic campaign run by a midsize retailer, Lucas explains how the Interactive Advertising Bureau's standard of 50% pixels for 1 second can declare an ad 'viewable' even when no human is paying attention. They explore research fro...

Episode 35 of Marketing Analytics with Fexingo dives into the underappreciated power of ad frequency caps. Lucas and Luna explore how setting the right frequency limit can dramatically improve campaign ROI, reduce ad fatigue, and prevent brand harm. They walk through real data from a 2025 e-commerce case where capping frequency at three impressions per user per week lifted conversion rates by 18 p...

Episode 34 of Marketing Analytics with Fexingo dives into a common but dangerous pitfall: multi-touch attribution models that look sophisticated but actually misallocate credit and inflate ROI. Lucas and Luna unpack a real-world example from a DTC brand that saw a 40% discrepancy between its attribution model and a simple holdout test. They explain why last-click models often outperform MTA in vol...

In this episode, Lucas and Luna dive into how multi-touch attribution directly affects a company's profit and loss statement. They explore a real case from a mid-size e-commerce brand that shifted from last-click to a custom weighted attribution model, resulting in a 15% increase in return on ad spend over six months. The hosts break down why attribution is more than just a reporting tool—it's a f...