All the data and analytics insights for analytics leaders we've scooped up this week - 9/2/22.
Analytics Leader Digest - 9/9/2022
Insights for analytics leaders. Data-driven leadership, evaluating your data stack, data analysis questions you should never ask and more.
What you need for data-driven leadership
Fail Fast, Learn Faster by Randy Bean
A book review of Fail Fast, Learn Faster: Lessons in Data-Driven Leadership in an Age of Disruption, Big Data, and AI by Randy Bean. This article summarizes really interesting points and survey stats on analytics and data transformation.
One interesting finding: Only 8% of surveyed executives felt that technology was the principal challenge in data transformation for their organization. The real challenges cited? People, process, and culture.
Is your company's data stack too brittle?
Nick Handel, CEO and Co-Founder at Transform, argues that most data stacks aren’t built for change, but there's a big opportunity that lies ahead. Dive in for a preview of how our data warehouses can become as dynamic as our organizations’ data needs.
If you aren’t familiar with the term “data stack”, its a loosely-coupled set of tools that create an end-to-end data solution without building everything from scratch. This post skips any conceptual background (you should know what a “DAG” is too) but any amount of pre-reading is worth it for this one.
Bad data analysis questions I see every week (and how to fix them)
Nothing sets an analyst up for failure like being asked the wrong question. Whether you’re the doomed analyst or the question asker (or both depending on the day), it’s useful to know what’s a “good” data question and what’s not. Working through a series of bad examples, this post gives tips on how to improve each. You’re likely to recognize something from your own experience. One insight that particularly resonated with me: beware the context gap between asker and analyst.
No silver bullets for software (or data) engineering
A thoughtful exploration of the path towards order-of-magnitude level improvements to efficiency in engineering projects. Written in 1987, some of Brooks’ examples have become outdated, but others like AI are still very much relevant. His core premise remains intact: separating complexity into the "essential" and the "accidental" continues to be a useful framework. For leaders attempting to slay the “werewolf” of software and data development, this is a good read.
The worst AI advice to heed
Like any poorly-understood, highly-coveted technology, AI has more than its share of snake oil salesmen. Folks are willing to say anything so long as it sounds good enough to sell. Out of the muck, this author claims to have identified the worst advice: “move fast, and break things”. Forgoing the Silicon Valley mantra, this article encourages leaders of AI projects to take a more deliberate, incremental approach. He argues that moving too fast in this still-nascent arena can lead to disaster.
Enjoy this roundup?
Subscribe to our weekly newsletter and get fresh insights for analytics leaders delivered straight to your inbox.