Closest I came was working with Schering-Plough (when it existed) and I made an interactive reporting thing out of Excel and VBA that was push-button by the time I was done and had to run weekly.
I pulled together SO MANY data sources.
Most fascinating to me was that I had data down to the Doctor/Patient (Insurance Plan) / Prescription level. 8000 Sales reps KNEW weekly before they came in, what was the most common insurance plans of a doctor’s patients and what that doctor liked to prescribe.
Honestly, it all felt borderline “invasion of privacy” but it was amazing.
If you could legally write a prescription in the USA, I knew your name
I absolutely HATED realignments though. Insurance companies would buy out each other, modify territories and names. And there was no rolling up insurance plans.
[Just because territory 9403 said Blue Cross of [place a], it had NO connection that was any use of me to territory 9409 that said Blue Cross of [place b] ]
From a data POV, it was annoying. I didn’t care about why — I’m sure it was to get stronger negotiating positions and profit of course – but to me I had to go in and get rid of a perfectly good name that lined up to zip codes, add a NEW name that lined up to OTHER zip codes including the originals… then change all of my ranges and stuff.
They spend $2 million on Microstrategy to replace me. After 9 months of working on it without consulting me, they had their demo. I wanted to laugh when I saw their “web based solution” that required the data person to hit “Browse…” for EVERY data file. Shit, I was pulling it right from mainframes and web and wherever straight into Excel. Idiots.
There’s about 3000 counties in the USA. I used to know how many zip codes. I actually used that info daily.
I don’t think in terms of states as there’s too many entities that operate cross state or are zip code or county dependent.