Discover more from Sanity Check
Analyzing in Public
Sanity Check • No. 013
Here’s what’s new:
👩🏫 For many it’s back-to-school time. For me, I’ve joined the CoRise - I mean Uplimit - Advanced dbt Course. The course has reinforced what I already knew and exposed me to new approaches, like Elementary for data quality monitoring.
💇🏻♂️ I couldn’t take the heat anymore. I got a summer chop. Not military short, but it’s no longer what my brother-in-law would call a “flow bucket.”
💧 Another brother-in-law introduced me to Topo-Chico. I’ve never understood why waiters ask “sparkling or still?” Now I have one reason to take the sparkling option.
A few interesting articles, podcasts, or websites I recently came across
I have been saying the secretive nature of internal analyses are a barrier to growing the data profession. I want to prove myself wrong.
Here are some of the best public analysis guides I’ve found:
SaaS Metrics 2.0 & definitions - this blog is canon for the SaaS B2B space. At ConnectWise this blog was our guide to measuring the growth of the business. That ability to measure, then identify the levers to influence the measures, is what enabled the company to have a unicorn exit.
SOMA B2B SaaS & repo - this Standard Operating Metrics & Analytics (SOMA) project is what I envision as modular analytics. The project includes fake source data and built-in reporting. To deploy at your company you only need to repoint the data feed from the fake source to your source systems.
Drilling Down - CPG/eCommerce companies will find this book helpful. While defined in spreadsheets, the work could be lifted & shifted to SQL. I operationalized the Recency-Frequency-Monetary (RFM) analysis in chapter 21 to help marekting identify customers primed for an upsell.
Kaggle - The ML ecosystem has done a much better job about being open with their projects. Kaggle is a great example of this openness. While the site is geared towards ML competitions, the datasets section is great for finding interesting business problems and the real-world messiness of datasets. A friend pointed me to WalMart’s retail sales as one example of high impact, but difficult to wrangle data.
Reforge Artifacts - Reforge’s programs are great, but now “Artifacts” highlights applied work examples - like Pinterest’s growth model. It’s helpful to see what is actually used to inform decisions at other companies. It is often more simplistic than you’d assume.
SQL for Data Analysis - This O’Reilly book delivers on implementation. It provides real datasets for you to practice applying the SQL alongside the reading. Time-series and cohort anlayses are covered in depth. Common gotchas are addressed - like `sum(sales) over (rows between 6 preceding and current row)` does not necessarily mean the same thing as “total sales this past week.”
What are some sources that you reference for your analyses? The more industry-specific the better!
Documenting analytics jargon, visually
Join me as I play a round of Squarely. If it looks fun, check out my dad’s work on Amazon 😊