When Being Right May be Wrong.

Recently I was tasked with automating a reporting process which collects repair metrics. The current method of tracking repairs in my company was tedious and time consuming. As is typical, the method used to track this was Excel Hell. Every week someone would run the Received Goods (RPREGO), Inventory On-Hand by Location (RPINOH), and the Monthly Shipment Analysis (RPMSHA) reports. They would export them to Excel, pivot table the information, and then type all of that information into a master spreadsheet which created multiple graphs for executive review.

Why was Monthly Shipment Analysis used rather than Invoiced Sales (RPIVSL)? Well, they want their information by product class and RPIVSL doesn’t pull that information.

Anyway, when I substituted an invoicing script instead, I matched their overall totals for repair revenue to the penny. However, the average repair cost on the invoicing script was quite a bit higher than they received from the Monthly Shipment Analysis. Upon investigation of our business processes I determined the cause. Often when a part is returned to us for repair, there is nothing wrong with it. Sometimes we will test the units gratis, and if no fault is found, return them to the company free of charge. These units are shipped, but no invoice is created because the dollar value is zero. When figuring averages their method has a greater quantity of repairs than mine (because it is derived from shippers, not invoices), so their average is less because those repairs are for $0. Make sense?

Luckily the manager acknowledged that by their own definition these units were not repaired, and so they went with the invoicing numbers with a footnote as to why the numbers were not consistent with statistics. However, the manager risked giving the executives the impression that his reports could not be trusted.

What do you think? Is consistency more important than accuracy?

My next article will discuss the ethics problems facing Business Intelligence Experts.

5 comments to When Being Right May be Wrong.

  • rk

    Accuracy is king, period. Just because an organization has historically asked the wrong question doesn’t mean it shouldn’t have access to the right answer if someone can provide it.

  • I’ll second rk on that one.

    It’s better to report nothing at all than something you know is wrong.

  • Kim

    Make it 3-0. I agree. I wouldn’t want to work somewhere that knowingly uses faulty information.

  • […] When Being Right May be Wrong. […]

  • Stephen

    We’ve got a phone system that tracks call stats. But the problem is, it’s set up wrong. We’ve suggested ways of making the report accurately reflect our phone activity, but I was told that the new reports wouldn’t be comparable to the previous ones.

    My answer to that was, “Well duh. These are accurate. The existing ones aren’t. Plus you guys are already complaining that the existing reports are inaccurate anyway! So you’ve got two choices: Go with accurate reports, or stop complaining.”

    They went with… Stop Complaining.

    Works for me! 🙂

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>