Skip to content

3 Years of Data Isn’t Enough: The Shocking Truth About Marketing Mix Models!

A company with an annual marketing budget of $250 million and revenue between $1.5 and $2 billion attempted to use Marketing Mix Models (MMMs) to optimize their marketing spend. The models were based on just three years of data, which equates to only 156 weeks. This amount of data proved insufficient when trying to account for various marketing channels such as TV, radio, press, out-of-home advertising, direct mail, email, search, and social media, along with prices and seasonal variables. The process also required adstocking each media to account for lags, which necessitated experimentation. The limited data made it challenging to create test and validation sets and to analyze the effects of combining different marketing channels. Despite the challenges, the company began implementing A/B testing in areas where the initial models indicated low returns, leading to new insights in previously untested areas.

Source: www.reddit.com

Related Videos

Related X Posts

Salif Sibane @Salifsibane16 · Jan 28
Quarter million in rev and 3k+ sales from this adThis creative is a masterclass to sell to a problem-aware audienceWe addressed the awareness level of the audience from the hook and throughout the lead, w/o revealing the offerTook us almost 30s to get to the productIf you

“Sudden And Unexpected” @toobaffled · Mar 26
Japanese researchers, free from the shackles of Big Pharma, have cracked open a mystery the world’s been dying to ignore—literally.The data is out there, the bodies are piling up, and the question burns: who’s hiding this, and how long until it blows sky-high?

BrainDo @braindo · Apr 10
To move people, your data needs proof, clarity, and insight. When done right, it doesn’t just inform – it inspires action.From the Marketing Roundtable podcast #DataStorytelling #MarketingInsights #AnalyticsWithImpact #MarketingRoundtable

Rod D. Martin @RodDMartin · Apr 3
4/ But here’s where it gets WILD: Critics have been saying for YEARS that the data was manipulated. The methodology? Flawed. The algorithms? They’d give you a hockey stick shape NO MATTER WHAT numbers you fed them!

BobbyGiggz @BobbyGiggz · Apr 10
http://medium.com/@BobbyGiggz/refining-data-why-meaning-and-context-are-the-real-products-of-the-information-economy-7c38744bda02… … “Refining Data: Why Meaning and Context Are the Real Products of the Information Economy”Summary Statement: In a world overflowing with raw data, true value is no longer found in access alone—but in the ability to interpret, contextualize, and

Jef Van Gool @JefVanGool · Apr 10
“We don’t need more leads — we need better ones.”A few weeks ago, I shared a post about how smart data analysis + a few tweaks turned a reasonable campaign into one that delivered 200% more qualified leads.Someone saw it. That someone had worked with us years ago — back when