When I look at Jupyter notebooks, I ask myself why is anyone bother with ootb reporting tools or growing their own? Jupyter notebooks are the right mix of customizable and self-documenting.
The kinds of analysis and visualizations you can build in 5 minutes with Tableau (and the ability to explore the space of possible analyses and visualizations) would take hours of futzing to reproduce with python.
You can build a basic report in 5 minutes. Then you’ll spend hours tweaking it and making all the changes the boss wants. And then hours more discovering there’s things you simply can’t do (but you’re boss won’t accept that).
And next week you will have to do it all again, because it’s all manual.
How many business users are going to use Juypter notebooks? Meanwhile, someone with basic computing experience can create in-depth reports with Tableau in less than an hour.
> How many business users are going to use Juypter notebooks?
None, because it's too much programming for IT to let business people have access to it, and it's not disguised as an office productivity app the way Excel is.
If they had access to it and had basic training on it that anyone already competent in any vaguely quantitative domain could handle, plenty of them could and would.
At least judging by my experience with SQL shells and similar told that are both less powerful and less friendly than Jupyter + Python, and yet plenty of business people used them productively in enterprise environments (often right up until IT ripped it from their hands.)
Python is making some inroads in shops that have been using SAS and/or R, but it’s a hard sell. Jupyter is even harder to push because of the server element.
SAS and R are pretty common business user tools that are basically the same thing as Juypter. Tableau is nice, but you generally need to use something else to prep data for it whether that's an ETL process IT sets up or something manual that an analyst publishes.
