Analytics Power Hour

On our latest episode, Christopher Dalla Riva joined to discuss his manual creation of a dataset with 1,100+ rows and 105 columns by listening to every #1 song on the Billboard Hot 100 going back to the list's inception in 1958. He put that dataset to use in multiple ways, including publishing a weekly Substack and writing a book! The parallels to more "typical" analytics work were many: how much of our data is simply a proxy for what we *really* want to know; how data can start out pretty simply and then grow in complexity out of necessity; how some questions require gathering still *more* data to meaningfully answer them; and how some questions, no matter how interesting or actionable, require data that is impossible to gather to answer! We think there were shoutouts to at least one tune from every decade going back to the 60s in the show. We would know for sure if we gave the show a really close listen and then did some supplemental research to compile a small dataset to make that claim definitively. But we're not going to do that. This episode's Measurement Bite from show sponsor Recast is an explanation of the miracle of randomization and how it addresses unobserved confounders from Michael Kaminsky! https://lnkd.in/gzKFf2MP #analytics #podcast #music #billboardhot100

What does “great” look like in data & analytics when you’re sitting in the CEO’s chair? On our latest episode, Flybuys CEO Anna Lee shared a candid C-suite perspective on: - Productive curiosity vs. chasing rabbit holes - Informed pragmatism: when “good enough” is the right answer - How analysts earn trust with leadership - The all-important “so what” every data team must deliver For anyone looking to influence strategy with evidence, this conversation is packed with lessons from the top down! This episode's Measurement Bite from show sponsor Recast is an overview of the fundamental problem of causal inference from Michael Kaminsky! https://lnkd.in/g7gnsEAz #DataCulture #AnalyticsLeadership #podcast #evidence #uncertainty #dataled

What is a dashboard? That was the title of episode #002 of this very podcast wayyy back in 2015. We are still awaiting the arrival of our hazmat suits to go back and give that a re-listen. According to the guest on our latest episode, Andy Cotgreave, and his co-authors (Amanda Makulec, Jeffrey Shaffer, and Steve Wexler) of the imminently releasing book, "Dashboards That Deliver: How to Design, Develop, and Deploy Dashboards That Work)", a dashboard is simply "a visual display of data used to monitor conditions and/or facilitate understanding." That definition got a LOT longer before it got trimmed back, and there are multiple shoutouts to Nick Desbarats and his categorization of different types of dashboards as context for that concise definition. Our latest episode was spawned by any number of thought pieces and musings that, like Most Everything, AI will soon bring the final demise of the dashboard by way of various question-answering chatbots and [mumbles something about "agentic AI"]. Not so fast? We covered a lot of ground—including a history lesson going back to the 1800s and a prescient quote from a book published in 1914! This episode's Measurement Bite from show sponsor Recast is a quick explanation of power analysis from Michael Kaminsky! https://lnkd.in/gSfRTG_K #analytics #podcast #dashboards #dashboardsmustdie #longlivedashboards #hashtagsmustdashboard

*Very* productive monthly production meeting at the podcast this month. Farmer Brown had some strong ideas about upcoming episodes.

Where does "process" fit into the role of an analyst? Arguably, it's at the core of how they can both consistently deliver business value while maintaining their own sanity. Yet, some analysts bristle at the thought—they've lived in overly process-heavy organizations where the word seems more like a synonym for "cumbersome bureaucracy" than a path to "consistent productivity". We tackled the subject on our latest episode. No guest for this one, but five co-hosts who are pretty well distributed across the process-enthusiam continuum, ranging from at least one (maybe two) Hearty Endorser to at least one (maybe two) Full-Throated Skeptic. Were we able to meet somewhere in the Happy Middle? You be the judge! https://lnkd.in/g8HYisSP #analytics #podcast #process

AI has certainly made these interesting times, hasn't it? In the world of analytics, AI can be the analyst's best friend when it comes to banging out SQL, Python, and regex. But is that best friend simply cozying up to the keyboard so that it can spell the end of the analyst's role entirely by becoming the business user's best friend instead—taking in business-friendly prompts and returning actionable insights? Technology platforms imply this is the case. Thought leaders are all over the place (but are very confident that they are correct). It seemed like a good time to sit down with Juliana J. to get her thoughts. She's been digging pretty deeply into different aspects of the topic for a while now, and she's got thoughts! She is also prone to well-deserved tirades about bold claims that are misguided! https://lnkd.in/g6fWA4f5 #analytics #ai #llms #podcast

Statistics is loaded with terminology and techniques that can feel pretty opaque. The funny thing is that, the more you dig into them, the more you find yourself bumping up against a relatively small set of foundational concepts. On our latest episode, we dove into the ANOVA waters with Chelsea Parlett to see where that took us. And we emerged in one piece feeling refreshed and just a little bit smarter from the experience! https://lnkd.in/gEBpdN3j #statistics #analytics #podcast #anova #ancova #manova

Are you excited that AI is finally—FINALLY!—going to add the capabilities to your company's BI platform that will make it the go-to self-service, easy-to-use mechanism by which everyone in the organization can safely and reliably access the data they need to generate actionable insights? Good luck with that. Colin Zima from Omni joined us for a lively discussion (the explicit tag on this episode is for real; Tim Wilson got a bit, er, animated) about BI platforms: a class of tools that seems to perpetually be overpromising and underwhelming. Why is that? Is it a limitation of the tools, an unrealistic and overly simplistic set of expections? Or both. https://lnkd.in/d78Wp3tB #bi #businessintelligence #bitools #analytics #dashboards #podcast

Look at you: perusing LinkedIn! Are you on the market for your next role? If not, chances are you will be at some point in the next few years. And, while we find it hard to believe, this particular social network is good for a lot more than just getting notified of new releases of your favorite explicit analytics podcast. Oh, and resumes still matter. As does human connection. Our latest episode with Albert Bellamy has some great perspectives and tips whether you're actively conducting a job search now or whether you likely will be in the next few years. https://lnkd.in/ghXDkTZw #analytics #jobsearch #podcast #recruiting #networking #resumes

On our latest episode, we had a very REAL discussion about synthetic data! We explored how it's made, where it's most useful, and some of the key considerations. We were joined by Winston Li, Founder at Arima for the discussion! https://lnkd.in/gysk2sEP #podcast #analytics #syntheticdata #datascience