I am having the following issue and would like to get some indication how you would solve such a problem.
We are sending parcels for customers. Some smaller customers send each second day e. g. 3 parcels, others every day 1 parcel and big customers send each day up to 100 parcels.
Now I would like to monitor the trends of each customers. Is he sending more or less compared to the last days / weeks?
When I calculate a percentage between the last days and show them wirh a graph this is pretty fine for big customers. They are stable with their volumes and you can identify the trend by looking onto the chart,
My issue are the small customers. Here by calculating the trend, of course sometimes it goes up to 300%, or the trend decreases by 200%, etc. Bringing them together into a big chart with the rest of customers, you can just see the big trends going up / down. And if you drill down to them, you identify that this is a very small customer with 10 parcels / month. So not really worth to monitor them.
How to you solve such problems? Which diagrams are you using?
I've been wrestling with this, and have decided to set charts so that the user can then control their view. I include a Timeframe set of buttons so users can view the last 7 days, 14 days, MTD, YTD, last 12 months and All time, with the dimension changing from day to month to year as appropriate (driven by variables). If filtering on a customer shows too widespread/little data on the week view, the user can manually switch to MTD, YTD etc.
Using a combo chart, current values are shown as a bar, and last week, last month, last year values as lines (droping last week and month when not relevant). A Compare To set of buttons can then colour-code the current bars to show whether they are better or worse than the last week/month/year.
It's tricky to set up, but can be applied to most areas off functionality.