Wednesday, January 08, 2014

ERP vs End User Computing

Today I had what might be considered to be a classic case in the continuing battle between using the standard tools of ERP as opposed to "rolling one's own" (end user computing, aka EUC).

Someone needed to give costs for parts which we have purchased in the past. As these parts are important, their costs are comprised of their ex-factory cost and the cost of shipping. Instead of trying to determine how much should be added for each shipment, we simply add a percentage of the price - in other words, if something costs 100 Euro EXF and its shipping cost is 20%, then we enter its cost as 120 Euro. We had reason to suspect that the shipping cost percentage for several parts had somehow been deleted, but we have no way of knowing when that percentage was deleted.

Had I been anyone else, I would have taken matters into my own hands and 'solved' the problem with my own insights and Excel (ie EUC). I would have looked at every part, determined when we last purchased it, looked at the delivery note for that purchase and derive from it the EXF cost, the exchange rate and the cost at which it was listed. If the listed cost is equal to the EXF times the exchange rate, then no shipping costs would have been added. This would take me a few minutes for each part.

But after a short walk which I had taken to exercise my muscles (I had been sitting on my chair for a few hours without moving), I thought a little about this problem. Surely it would be better to get a professional programmer (myself, under a different hat) to write a program which would give the required data. The advantages of this method are that the data are guaranteed to be correct and that the time involved would be fixed - in other words, I could devote 30 minutes to writing the program and this program would be available for whoever needed it, whenever (as it happens, I think it took me even less than 30 minutes to write the program, including the necessary debugging).

If I had fifteen parts to check now and every part took two minutes to check, then time-wise I would come out even. But if I have another ten parts to check, the cost in time to check them would be zero. And if I had three thousand parts to check ....

So why do people prefer to find their own solutions, even if they are not efficient? I can think of a few reasons
  • they never consider transferring the problem to a professional
  • they are unable to explain the problem to said professional
  • they think that passing a problem on to someone else to solve somehow makes them a less good worker (because they didn't solve the problem themself)
  • they think that they can get the data by themself in a few minutes whereas sending the problem to a professional will mean that the solution will take more time
Although this isn't exactly what I want to research (I am more interested at the moment in researching the extent of EUC in companies running ERP), it may well be that I will also have to research why people do this. I may use the above as a test case and try and find the real reasons why people use EUC. The problem may be that users aren't necessarily able to explain their actions - not everybody thinks about these abstract ideas like I do (in fact, I doubt whether anyone thinks about them at all!). 

[SO: 3282;2,12,30]

No comments: