Saturday, February 13, 2016

ERP thoughts

Despite today being Saturday, my internal alarm clock woke me as usual at 5:30am. Also as usual, my mind was filled with abstract thoughts, initially about a question which had been asked on the 'English Language Learners' stack exchange site about the difference between 'I forgot' and 'I forget'. From there, it was only a short leap to thinking about my DBA submission (I suppose I should get used to calling it a thesis).

A few days ago, I wrote on a DBA forum that one should include - as I neglected to do - answers to the following questions:
  • What is this research about?
  • What does it hope to achieve?
  • What is its value?
From there, my mind started dealing with an expression which I sometimes use: ERP deals with the past whereas spreadsheets deal with the future. This sounds quite good, but it's not entirely accurate: ERP can deal with the future, whereas spreadsheets can deal with the conditional, or as I learned in Latin grammar so many years ago, the subjunctive.

Fired up, I went to the computer - this is at 5:45 am, before taking the dog for a walk - and wrote the following. It probably needs a little polishing for the thesis but it's clear enough for here. It's also a fragment and taken slightly out of context, so it may seem incomplete.

ERP systems excel at answering the question “What happened?”. They are also good at answering the question “What needs to be done?” (by customer, purchase and work orders). But they fail to answer the question “What if …”.

One reason for this is that ERP systems are fixed in terms of their programming and parameters, and it is unlikely that the programmer could be able to predict every parameter which a user would wish to change. An example: the researcher wrote a report which shows the cost of a manufactured product according to its bill of materials; the report shows the cost at which the constituent parts were purchased in the past. In order to answer the question “What if there is a change in the dollar or euro rate of exchange?”, parameters were added to the report allowing the user to define fictional rates of exchange, thus allowing the question to be answered.

But the report would not be able to answer the question “How much would this product cost if we used 10% fewer widgets, or reduced overhead by 15% overhead, or …”; such a question is open ended as the programmer cannot foresee which parameters need to be changed. The report could be changed ad hoc but this would require programming skills and also means that report runs could not be duplicated. Thus ‘what if’ questioning is traditionally found in the realm of EUC [End user computing, normally meaning spreadsheets].

No comments: