Building a causal insights program that actually changes go-to-market decisions.
Madan Bharadwaj, Lydia Pittman, and Andrew Huang on turning causal measurement into real go-to-market decisions.
At eTail East in Boston, M-Squared founder Madan Bharadwaj was joined on stage by Lydia Pittman (VP, Direct-to-Consumer Sales, Virginia Food Group) and Andrew Huang (VP, Marketing, Insurify). The session opened with a crash course in causal attribution and why last-click and platform self-attribution systematically mislead, followed by a blueprint for shaping a causal insights program: a repeatable
system that stacks a Marketing Accounting Framework, marketing mix modeling, geospatial analysis, and incrementality testing so insights compound on each other.
The heart of the session was two practitioner case studies. Lydia walked through how Virginia Food Group is using advanced measurement to turn a seasonal gifting brand into an all-season habit, and Andrew showed how Insurify rebuilt its marketing truth, moving from a lower-funnel article of faith to a tested, causal case for upper-funnel scale. The session closed with a fireside chat and audience Q&A.
Eight ideas worth keeping.
Attribution hasn't kept up with the consumer journey.
Journeys are now cross-channel, but attribution is still mostly last-click. Site analytics over-credits lower-funnel channels like branded search, and media platforms' self-attributed conversions typically add up to more than the conversions that actually happened.
Causal attribution is the answer.
Marketing mix modeling and incrementality testing, combined with attribution tracking, provide the signal needed to make efficient investment decisions. These are expressed as causal multipliers on last-click reporting that vary wildly by channel, from around 20% for brand search to more than 30x for podcast and broadcast.
A Marketing Accounting Framework is the first step.
Not all revenue is equal, and aggregated results blur the truth. Virginia Food Group was broken into 11 P&Ls across three brands, and Insurify's quote volume into three customer segments. Each behaves differently and each gets its own model.
A causal insights program is a repeatable system.
Stack analytical techniques one step at a time: MAF, MMM, geospatial analysis or interaction sub-models, then incrementality tests. Each insight compounds on the last and feeds a governance cadence of QBRs, refreshes, and budget guidance.
DTC case study: the everyday buyer isn't where the gifting buyer is.
VFG's MMMs showed category demand, not media, drives most orders. Stripping out gifting weeks and rebuilding on the everyday signal revealed a different buyer, one defined by income, home value, and education, along with thousands of greenfield ZIPs now being validated with a publisher-partnered CTV incrementality test.
Lead-gen case study: interrogate the model, don't defend it.
Insurify's first MMM cut showed CTV and Linear TV below breakeven. Re-specifying TV explicitly as a driver of lower-funnel performance moved CTV from 0.8 to 1.5 and Linear TV from 0.8 to 1.2 MMM ROAS. That is real evidence that upper funnel drives lower-funnel efficiency.
Prove it before you scale it.
An insight is a hypothesis until it survives a live test. Both brands funded incrementality tests before committing the budget: a CTV geo test in flagged ZIPs for VFG, and a CTV holdout, a Linear TV natural experiment, and a Paid Social lift study for Insurify.
Real measurement is a partnership.
It's about understanding the marketing leader's go-to-market thesis and finding alpha that calibrates that thesis with data. Models produce reads; expert interpretation turns them into decisions the business can act on.
The full eTail East deck.
Every framework, chart, and case study from the session, exactly as presented.
About M‑Squared.
M‑Squared is a finance-grade marketing measurement and incrementality partner, built for marketers who want to understand what is actually driving results.
Talk to an expert.
Walk through your measurement program, your media mix, and where a causal insights program would move the decision.