When we talk to people who are new to MMM and tell them there’s no pixel and we don’t track people at all, we basically see their minds explode. The idea of trying to track people as well as possible has become so established that it can be hard to imagine how this could possibly …
Understanding the Enigma of Marketing Mix Modeling (MMM) without Tracking Read More »
To prepare for the future, we must understand the past. This stays true in marketing, and it is particularly appropriate when it comes to digital tracking and analytics. D2C/ecommerce brands have often trusted their attribution and budget allocation decision-making on tracking people across the internet – but that has changed as privacy concerns and legislation …
Coca-Cola does not think about attribution the same way an early-stage D2C brand thinks about it. Everything is different: budgets, teams, brand recognition, channels, etc. Both are considering which initiatives are incremental, but how they go about it is an entirely different game. Attribution is not a one-size-fits-all method, and each marketer needs to understand …
How Marketers Should Think About Attribution at Different Stages of Growth Read More »
In the world of website analytics, the concept of “dark social” has emerged as a significant challenge for accurate channel/source reporting. It refers to web traffic that is incorrectly attributed as “direct” in analytics when, in reality, it is driven by social networks and other hidden referral sources. This phenomenon has far-reaching implications for marketers …
Google Analytics has recently gone through significant changes that have sparked concerns among marketers around its bias and validity. As digital tracking continues to break, we need the right models and tools that will help us make the best decisions around budget allocation and, because of that, we need to question what the role of …
Your marketing performance doesn’t live in a silo and it’s not independent of external factors. One of those that will impact it is GDP (gross domestic product). But that brings many questions to marketers: If our marketing is doing well during a period of GDP growth, is it because our marketing is effective or is …
Why You Need to Stop Controlling for GDP Growth (and what to do instead) Read More »
With privacy concerns on the rise and digital tracking becoming less reliable, marketers need advanced tools to measure the impact of their campaigns. Geo-testing has become an alternative used by many brands (often in conjunction with MMM) since it uses aggregate data and doesn’t depend on following people across the Internet. Geo-testing is gaining the …
Should You Be Using the GeoLift Library to Run Geo-Testing Lift Tests? Read More »
Heteroskedasticity in a regression model is when the variance of the residuals varies with each observation. Most statistical software assumes that the errors made in prediction are random noise, or in statistics speak, the variance of the residuals is constant for all observations, which is called homoskedasticity. Heteroskedasticity, where there’s a pattern in the errors …
How to fix heteroskedasticity in your Marketing Mix Model Read More »
Brands have been using models to understand the effectiveness of their marketing activities since the ‘Golden Era’ of advertising in the 1960s. However, as our history of Econometrics details, the dawn of the internet era and digital marketing in the early 2000s moved Marketing Mix Modeling (MMM) out of the hands of the select few …
Any consultant or analyst has projects which they regard as great successes, and will gladly talk about at length, maybe even writing case studies about them. But they will also have projects which went less well, and this is often where the best lessons are to be found. One of the worst modelling projects that …
