In business analytics backend > frontend. If you are a business analyst you want all your data at your finger tips + the ability of running somewhat complex transformations and calculations. You also need to be able to validate data with your own rules, because you can't always trust data from warehouses and other sources. Then you want to take all your work and schedule it, so it runs automatically and refreshes all the reports depending on it. Finally, you want triggers on data changes/updates and good looking email notifications with screenshots, attachments, etc.. Of all of the above, end users might only see the dashboards, pdf's or Excel's on the frontend part. If you have a solid backend, you will be able to serve your customers at multiple times the speed and with a lot more insightful reports. At that point the difference between a Tableau dashboard (which does look good) and, say, Superset (which does not look as good) is somewhat negligible.