The Power of Carefully Worded Nonsense

A model is useful if it describes something in a context that renders a complex topic easier to digest for specific audience. Its purpose is to describe, to communicate. It is an expression of a viewpoint. It isn't a detailed map or a blueprint. It is representative of a system, it doesn't depict the system. It is an abstraction. I'm afraid we model too frequently as a cover for not actually producing things of value.

There was a great article written about a week ago that touched so many nerves it spawned a whole subject area on the blogger’s site.

The basic premise was just how ridiculous some of the thinking and positioning around social media had become. I urge you to read the entire thing and navigate the thread of follow-on posts.

This immediately stuck me as being particularly relevant not just to IT in general, but architecture and strategy specifically. I think this is primarily because any field or topic associated with or practiced by consultants is rather susceptible to using ridiculous language and outright hyperbole from time to time.

I was in the midst of thinking through some absurdities when I was hit with a request for a reference model. It occurred to me, is it really necessary for us to model everything? Doesn’t it become a bit overdone at some point? I mean, frameworks, models, models of models and meta-models? Really? Why is IT engaged in a rush to model ourselves to death?

The Twitterwebs were all abuzz while I was thinking about this topic. Apparently there was some conference someplace with plenty of folks tweeting away. The ephemeral stream of wisdom included things such as “What’s the next big thing in EA” and “The need for better BPM models in EA”. Meanwhile I was being bombarded by emails from a colleague with model frameworks that we might potentially use for a project.

To be honest, I’ve had it with models.

A model is useful if it describes something in a context that renders a complex topic easier to digest for specific audience. Its purpose is to describe, to communicate. It is an expression of a viewpoint. It isn’t a detailed map or a blueprint. It is representative of a system, it doesn’t depict the system. It is an abstraction.

I’m afraid we model too frequently as a cover for not actually producing things of value. Maybe I’m being cynical again (there I go…!) I am merely observing, looking around at my field and taking note of the sheer number and variety of ways to describe essentially the same things. I’d be more impressed if someone presented their model and showed explicitly how it was used in a context that generated some sort of value (i.e. dollars). Haven’t seen that yet. In fact, I know folks who’ve been modeling the same system for over a year and I have yet to actually see any results from their work! I’m beginning to suspect this is much more common than I had already thought it was.

I recently had to endure a presentation on our latest model of BPM within a particular enterprise. Whereas I had proposed an approach that utilized exactly one slide with one set of graphics communicating one model, I was faced with an onslaught of thirty-two slides describing the counter-proposal. Each slide was essentially the same perspective, through a different lens. In one case, the model on one slide was literally the same model as on another, but this time it was ‘vertical’ instead of ‘horizontal’. Seriously?

At the end of that presentation, it became clear that the team of architects had spent two weeks working on models of the BPM approach instead of actually thinking through the BPM approach. From their enthusiasm and recommendation that we follow-up with alternate views, I have come to the conclusion that they preferred to model the solution rather than actually work on the solution.

In short, to recap in a slightly more succinct version, I think modeling is overused, misused and totally abused. It is becoming a proxy for wasting time and client money but with a more professional sounding name. To paraphrase the article linked above, it may be high time we start thinking about deploying Modeling Strategists.