Data Analyst: So what do you want to measure? What question do you want to answer?
Customer: Can you do a column chart, where I can see how many Orders we have?
Data Analyst: Column chart? What's the Axis? Per day?
Customer: No, per month.
Data Analyst: Right, so new Orders per month?
Customer: No, how many we have in general, new and old.
Data Analyst: Do you mean the old ones still open at the start of the month?
Customer: That's a good idea, yeah. Actually, can you add the ones we complete in that month too?
Data Analyst: The amount of completed orders? That would double-count them.

*shared moment of confusion*

Customer: Don't make it so complicated, I just want to see how many orders we had.
Data Analyst: Let me ask again, what question do you want to answer?
Customer: I want to know how much our teams are working.
Data Analyst: As in, how many orders they're completing?
Customer: I also want to see if we need more people.
Data Analyst: Like, if they can't complete all their orders? So basically, the rate of completed versus new ones?
Customer: Ooooh, good idea, can you put that rate as a line over our chart of new, old and completed orders?

Customer: Oh, and the warranty returns too! They need to be processed as well, that's also work.

Customer: Actually, we have this task tracking for who does which work for the order or warranty return.
Data Analyst: Shouldn't we use that to track how much work the teams are doing?
Customer: Yes, put it in the chart too.

Epilogue: The Customer got a separate chart for the tasks - turns out I'm not charging by the chart, so you don't need to cram as much as possible into a single chart. They also were persuaded to stick with "Old" and "New" to show the total workload, with the "Old" bars providing an indicator for how much stayed open and whether the backlog was growing.