Waiting for a connection and syncing reliably and testing that it works is a little harder -- annoying but not too bad.
Provisioning a database to store the data, maintain backups, index properly for queries etc., and maintain so it doesn't run out of space or degrade in performance as datasets grow by many orders of magnitude is an entirely separate skillset and harder still.
Visualizing it -- coming up with complex SQL queries and then dumping them into a graphing tool and overlapping different data series on the same graph -- will take forever. A big part of analytics is exploring and looking for patterns. The person with the business background doing the exploring probably isn't a programmer or a SQL expert. So a pre-built GUI for this is huge, and will be the hardest thing of all to program.
So, pretty hard.
Or... you can throw in a third-party tool with all of that, and have it all work in an hour or two.
But the last two steps can happen on your side. The data can go to a data store someone else maintains like AWS and you can use any number of packaged visualization tools on top of it.
Waiting for a connection and syncing reliably and testing that it works is a little harder -- annoying but not too bad.
Provisioning a database to store the data, maintain backups, index properly for queries etc., and maintain so it doesn't run out of space or degrade in performance as datasets grow by many orders of magnitude is an entirely separate skillset and harder still.
Visualizing it -- coming up with complex SQL queries and then dumping them into a graphing tool and overlapping different data series on the same graph -- will take forever. A big part of analytics is exploring and looking for patterns. The person with the business background doing the exploring probably isn't a programmer or a SQL expert. So a pre-built GUI for this is huge, and will be the hardest thing of all to program.
So, pretty hard.
Or... you can throw in a third-party tool with all of that, and have it all work in an hour or two.