≡ Menu

When is a cockpit not a cockpit?


When it’s a dashboard.

I often hear people talk about fighter pilots as a model for the future of business users. For instance, Ade McCormack did in his book on the IT value stack. Indeed this is such a good mental image that many companies use it – either to promote their products or to talk about how their internal IT department is going to change the business and support innovation. But go look at a cockpit (here’s a google search for you in case you don’t have a plane handy).

Done looking? What are the three main components?

  1. Things that tell you how the plane is doing, where it’s going etc – dials, gauges, lights
  2. A windshield so you can see out
  3. Ways to change what the plane is doing – knobs, switches, levers, etc.

To “fly” a business you need the same things – you need to know how it is doing, you need to be able to see the context in which it is operating and you need to have ways to change what is going on. Let’s test that analogy:

  • A performance management or event-centric dashboard gives you a pretty good view of how the business is doing (obviously reporting and traditional BI fails even this test but let’s keep moving here).
  • Most business people know their industry, keep track of the overall economy and news that might impact them – the business equivalent of looking out of the window.
  • What almost no business has is the set of knobs and levers to change the behavior of the business in response to either of these.

Think about it. If the dashboard tells you too many calls are being routed from the automated system to a person, how quickly could you change that? If the dashboard says that the website cross-sell rate is below industry averages, what would you do? if you hear that a competitor is going to start offering 0-down, 0% loans, what could you do about it? If the regulator’s change the way you should price something or the reserves you should keep, what could you do about it? The answer for all too many companies is “call IT and yell at them”. Not particularly constructive.

This is why a cockpit is a great analogy for the use of enterprise decision management. If you have used EDM to externalize and automate the critical operational decisions in your business, and exposed those decisions so that business users can work directly on them, then you can use EDM to “fly” a business – the exposed decision logic elements are the knobs and levers.

I once wrote that, when it comes to BI, if dashboards are the end game kill me now. So use EDM and build cockpits, not dashboards.


Comments on this entry are closed.

  • Tom J June 10, 2008, 3:22 pm

    Spot on with the analysis here. I’ll offer this as the litmus test for a dashboard metric; does the metric have a pre-defined action to take if the value goes above or below a predetermined value? If so, then it is a valid metric, if not then it is an irrelevant indicator and a waste of resources.

  • James Taylor June 11, 2008, 9:57 am

    Great point Tom – if you are not going to react to something it can’t possibly really matter!

  • Dan Keldsen July 1, 2008, 2:46 pm

    That which gets measured, gets done. Except when nobody acts on it! Ah, Yogi Berra where do you stand on EDM? 🙂
    Interesting thoughts – and @tomj where I thought you MIGHT be going is “are there auto-corrective measures that kick in once you pass high/low thresholds on your measures?” – the very least of which should be an alert that SOMEONE SHOULD LOOK AT THIS, rather than “Yes, our dashboard has been showing RED for some time now. Someone (who?) should do something (what?) about that someday (under what SLA, etc.?).”
    Well, no end to the work that COULD be done, and that is, by the smarter companies out there.