Tuesday, 18 May 2010

CoreITSM 101: Reporting

I think it was Horace who said 'I labour to be brief but become obscure"

How apt for this post. Firstly recognise that any reports we produce are an attempt to distil reality that is doomed to failure and secondly many reports lose their meaning in translation  from IT terminology to business speak, or BS as some of us call it. In fact as far as the business is concerned are reports might as well be latine scripta sunt in. And of  course what  Horace actually wrote was

"Brevis esse laboro, obscurus fio" 

At this point you might nod sagely and recall Churchill saying

"There are lies, damned lies, and statistics."  

Shame on you though if you think that means statistics are stuff and nonsense. If you knew your Churchill you would know the great value he placed on statistical information. It is no coincidence that operational research grew up as a discipline during WW2.

I spent a lot of my early career producing New Scotland Yard's crime statistics so I could bore you to death for hours but let me keep it very simple :
  • Bad reports are worthless
  • Good reports are priceless
What is a good report?Again, very simple: It tells you something you did not not know before reading it.

Now if I want to fill a conference session I'll speak about reporting . Whatever advice I give, l know that what most of the audience actually want is a standard  list of the metrics they should produce .

Which means they haven't listened to a word I've said.

There is no golden list. What you need to report on is driven  by context. A report that might hav been vital last year might be just so much noise this year.

I hate doing this, but sometimes I have to. Let me give you some guidance that will help you realise you aren't producing useful reports. If it makes you feel better these are all lessons I've learnt the hard way .
  • If you don't produce the figures one month, does anyone notice or care ?
  • Do managers just pin your graphs up to decorate their offices ?
  • Can you tell a story about what the figures are telling you?
    You should be able to write at least three sentences:
    • what the figures mean
    • should we worry about what the figures mean
    • what should we do in response to the figures
Of course of you are on top of your game you will have:
  • Looked at individual figures in the context of what other figured are telling you.
  • Identified in advance the thresholds for when action is required, and those thresholds will enable you to avoid failure, not respond to it
  • Thought through the actions people need to take.
So today's homework: go away and assess the reports you currently produce against these criteria.


  1. What methodology do you use to identify the metrics you should or would like to collect?

    Do you use something like GQM - starting first with the goals/purpose...

    Or something more like how ITIL seems to go with starting with the metrics that you want/can collect and seeing if they make sense...

  2. First of all my apologies for a tardy response. I hope I went someway to answer this question on my Brightalk webinar on metrics. The starting point needs to be what is important, not what is easy to collect, and the metrics have to be purposeful. Not only that but a metric that was useful last year might be nothing more than noise this year.