My data hot take is that the most data driven companies run their analyses in Excel in the beginning, the modern data stack be damned, while companies who start out with a modern stack tend to have over-engineered data departments serving execs who prefer to go with their gut.
Conversation
This is totally not backed by anything more than idle speculation, extrapolating from a very small sample size.
But … if the company is seeded with really good metrics-driven operational types, they tend to start out with Excel? Because they can’t wait for data infra?
Replying to
“Oh yeah, we figured out that these good outcomes are driven by these five leading indicators, so we have Joe in finance update them manually for us every week. Took us awhile to get there, though.”
“You have who do what how?!”
1
5
Arguably you have a very small window in time where this is true (poor Joe!), but I now think it bodes well if you have some process control or growth model running in a spreadsheet … before you build out all your data infra.
5
Replying to
From what I've seen, "modern data stack" is definitely overkill at the start
But folks who have been burned by Excel before (can have human IO errors, not real-time updated, no alerts) seem to prefer simple tools like Google Data Studio or Zapier+AirTable instead
Replying to
We make SQL driven automatic dashboards before we launch features, so that we can’t trick ourselves with clever human operations in excel. We make them deliberately not fancy though?
1
1
I do like though that is mandatory to make the query/dashboard before the product launch
1