Summary In today's episode, I walk through a data-driven method for deciding when to refresh or retire your ads instead of relying on your feelings. Here's what this means for you. You gain a clear, repeatable technique that detects diminishing returns in your ad performance so you can protect your budget and boost ROI. You'll…
Summary In today's episode, I announce the end of Saturday Night Data Party and explain why marketers must adapt to shrinking data access and increasingly unreliable research. Here's what this means for you. You gain a clear framework for becoming a data custodian and applying rigorous peer-review thinking to everyday marketing analysis. You'll also learn…
Summary In today's episode, I walk through how to chip away at building a data-driven SWOT analysis on a small team that lacks key data skills. Here's what this means for you. You gain a practical starting framework for turning limited resources into actionable competitive intelligence. You'll also learn these concepts: how to identify which…
Summary In today's episode, I explain how brands can use large-scale data analysis to figure out which Facebook ads actually work in their industry. Here's what this means for you. You can stop guessing about ad creative and instead mine competitor posts, tag them for themes, and test data-driven variations against your normal ads. You'll…
Dasha asks, “I want to take some classes on analytics and visualization skills – what would you recommend?” I’d start by learning the principles of data visualization first. Edward Tufte’s book, The Visualization of Quantitative Information, is the classic textbook to start with. Then look at Data Studio’s introductory course, followed by Microsoft’s free EdX…
Jessica asks, “What are the best tools for cleaning data?” That’s a fairly broad question. It’s heavily dependent on what the data is, but I can tell you one tool that will always be key to data cleansing no matter what data set. It’s the neural network between your ears. Can’t see anything? Watch it…
Summary In today's episode, I walk through the best tools for cleaning data and why the right choice depends on your data type, what you're trying to fix, and the scale of the problem. Here's what this means for you. You'll see why human judgment and process matter far more than any specific software tool.…
Summary In today's episode, I walk through a year-over-year analysis of FBI hate crime data comparing 2017 to 2018 and reveal what the numbers actually say versus what they appear to say. Here's what this means for you. You'll see how a metric can look great on paper while the underlying data quality quietly collapses…
Steve asks, “I’m looking for a dataset of companies that are actively selling on Amazon. How would you as a marketing data scientist approach this problem?” That’s an interesting question. To my knowledge, there aren’t publicly available, free datasets of this sort (though please leave a link in the comments if you know one), so…
Summary In today's episode, I walk through a practical methodology for building your own dataset of Amazon sellers when no public source exists. Here's what this means for you. You gain a reproducible framework for identifying e-commerce sellers using web technology signals, saving you from relying on unvetted vendor lists. You'll also learn these concepts:…
