Hi fellow data professional!

I had a very adult weekend after baby bedtime.

I uncorked a bottle of wine, cracked open my laptop and… stayed up late making a dashboard tracking my assets and debts.

While I learned some important high-level insights that will help me make financial decisions as I renovate and prepare to move into my home, I realized a basic development truth that could help anyone knee-deep in their own project.

If your goal is commit or ship then the simplest implementation wins, especially if you (like me) are perpetually time-constrained.

Simple execution doesn’t mean cutting corners. It does, however, require you to carefully consider any trade offs you need to make to get the thing done.

For instance, I originally wanted my financial dashboard to be connected to my data warehouse and fueled by production-grade ETL pipelines like I build at work.

But with only an hour or two to work on this on a single weekend, I opted to choose a data engineer’s frenemy: The dreaded spreadsheet.

I initially envisioned a cool, interactive dash with a slick UX inside which I could toggle metrics and do time series analysis.

I wanted to build super performant views in BigQuery so I could just drag/drop key metrics into the dashboard. These views eventually got downgraded to Looker calculated fields.

As I made these compromises, I realized I would only be using this for a “10,000 foot view” of my finances; I would be reading it in my mail app and wouldn’t touch Looker unless anything needed to be tweaked.

Given those constraints, I built for my target audience: Me—an information saturated, sleep-deprived dad. I threw out nearly all graphs because I didn’t feel like pinching and zooming.

I positioned key metrics like total monthly and annual interest amounts inside “score cards” so I could glance at a single, top-line number.

In the end I was happy with the final build because even though it wasn’t the shiny “command center” I envisioned, it was finished and actually fulfilled my use case.

When you find yourself with a bit of development block, I encourage you to reduce your ambitions to build a minimum viable product (MVP). I’ve found the act of building can reignite that excitement you initially had for a stale idea or project. Or, if the spark is gone, at least it’s done.

If you're struggling to complete your first data analytics project, steal this stack:

• Google Sheet
• Looker Studio (free to sign up)
• G-Suite (for email delivery)

Those three components are all you need to build an automated solution in a handful of hours.

And if you want a better idea of how to build an MVP dashboard, check out my final product.

Thanks for ingesting,

-Zach

​Medium | LinkedIn | Ebooks