Zach, with low literacy levels you can't say "this is good, this is bad", because people will not understand why. A reference model is not enough, you must guide them step by step, make them compare outputs and explain why a 3D pie chart is not good for them. This is not easy and takes a long time, but there is no other way, I think.

Jorge, check the link to Stephen Few's website. He's got a whole page of "bad graph" examples along with analysis and explanation of a proper solution. Pretty good reading so far.

I like the framework, but I think it's missing the cyclical relationship between these three activities:

1) Start with exploration/analysis, until you get to the point where the thing you're analyzing is "well-understood."

2) Develop some reports that help you identify patterns, trends, and notable exceptions.

3) Present the "insights and understanding" that you found along the way to your colleagues - this is bound to generate new questions, which brings us back to step 1.

This is admittedly oversimplified, but I think its a workable model.

I envision the diagram of this process as a circle with arrows showing the process flow, rather than as a triangle.

Rob, I know the page and it is a good starting point. This is one of the things I also try to do in my blog. But you need some sort of interaction, to explain, overcome objections, etc.

I would say that we are selling something that no one wants to buy. Imagine an ad were everyone is happily drinking and socializing. Then we must sell the message "don't drink and drive". People know you're right, but they'll keep drinking (and using 3D pie charts because the boss likes them).

Spot on identifying the problem, and I like the model you presented. The last mile challenge is probably more a people issue than any other. Exploration, understanding and presentation of insight require unique and uncommon skills, and talent. There is only so much that can be embedded and automated in a system. After that it comes down to the people that deliver the last mile.

Friends don't let friends build 3D pie charts.

Hahaha, nice one Dave!

What's worse than a pie chart? A 3D pie chart!

I'm surprised that you think of the "last mile" as a pervasive issue. Yes, it is true that several interfaces leave a lot to be desired in terms of content and usefulness, but I contend that leading corporations are very familiar with and extremely good at mitigating the risk of "lost in translation" between technical and management teams. With 15+ years of robust global business growth and the emergence of IT as key enabler, corporate success is in large part attributed to the availability and relevance of key business insight through the use of well designed EIS and BI systems. So I agree with your hypothesis in part but would say that it is really a matter of the experience level of the people involved. When you look at leading BI implementations, you will see tech and management working in concert to create value for the organization. It’s really no different from building any other kind of system.

I keeping thinking of Mr Covey when I deal with clients facing this issue. I believe it is either his 1st or 2nd principle that says, "begin with the end in mind." I can't help but think that the last mile is so challenging because we start with a list of requirements or features and not of desired outcomes.

We are in the "Last Mile" in my project where I am embarrassed to say that I manage the User Interface. I struggle with everything you have mentioned above. Unfortunately, the company I work for does not feel it important to invest in UI and my team is sparse and has been over run with the Data Modelers with no design sense. If just 2% of our time was spent on usability we would have much easier to understand product and the training to use it would be greatly reduced.. I will keep reading your blogs.