Using Visuals to Make Decisions
Many organizations, especially IT organizations, are extremely effective at collecting information. This isn't a good thing. The data comes in heaps, with no purpose or structure, and often gets dumped into an unimaginably complex and lengthy powerpoint presentation or built into a dashboard that is at best interesting and at worst shuts down the higher brain functions.
How can an executive cope with the deluge? I am a big fan of crisp, simple visuals. Avinash Kaushik over at Occam's Razor recently posted an excellent example of how a visual can make an enormous difference in making data decisionable. The post features a great graphic depicting the breakout of the federal budget.
Avinash's post is a recent example of what a few have known for a long time: proper visualization of data is critical and it's not easy. My favorite tome on the subject is Edward Tufte's The Visual Display of Quantitative Information. You can also locate Mr. Tufe's website here.
In my view, which some may view as counter-intuitive, the visual can be an excellent starting point for designing a decision making process and a data collection program to support it. If you are an executive, especially a visually thinking executive, try the following steps the next time you are facing a complex decision:
1) Set the decision goal - what's the outcome this decision is intended to resolve or advance?
2) Draw, by hand, a picture that would contain 80% of what you'd need to make the decision. It might contain the criteria, the data sources, the players, or any number of other influencing factors. Don't be afraid to sketch if you (like me) can't do much better than stick figures. Don't worry about accuracy.
3) Hand the picture to your corporate or organizational metrics or business intelligence POC. Tell him or her to turn around a 60% accurate representation of it within 24 hours. If they look puzzled, there's a good chance that either a) they don't know how to represent corporate data in a way that is decisionable or b) you are an exceptionally poor artist. "SAP doesn't present that report" is no excuse.
4) Iterate the picture until it's accurate enough to make the decision in question on an ongoing basis. It might be hard to get the first couple done, but keep at it. Once the organization has rewired its neural pathways to get the report done quickly, it should be a breeze to repeat.
How can an executive cope with the deluge? I am a big fan of crisp, simple visuals. Avinash Kaushik over at Occam's Razor recently posted an excellent example of how a visual can make an enormous difference in making data decisionable. The post features a great graphic depicting the breakout of the federal budget.
Avinash's post is a recent example of what a few have known for a long time: proper visualization of data is critical and it's not easy. My favorite tome on the subject is Edward Tufte's The Visual Display of Quantitative Information. You can also locate Mr. Tufe's website here.
In my view, which some may view as counter-intuitive, the visual can be an excellent starting point for designing a decision making process and a data collection program to support it. If you are an executive, especially a visually thinking executive, try the following steps the next time you are facing a complex decision:
1) Set the decision goal - what's the outcome this decision is intended to resolve or advance?
2) Draw, by hand, a picture that would contain 80% of what you'd need to make the decision. It might contain the criteria, the data sources, the players, or any number of other influencing factors. Don't be afraid to sketch if you (like me) can't do much better than stick figures. Don't worry about accuracy.
3) Hand the picture to your corporate or organizational metrics or business intelligence POC. Tell him or her to turn around a 60% accurate representation of it within 24 hours. If they look puzzled, there's a good chance that either a) they don't know how to represent corporate data in a way that is decisionable or b) you are an exceptionally poor artist. "SAP doesn't present that report" is no excuse.
4) Iterate the picture until it's accurate enough to make the decision in question on an ongoing basis. It might be hard to get the first couple done, but keep at it. Once the organization has rewired its neural pathways to get the report done quickly, it should be a breeze to repeat.
posted by John Dillard at
10:56 AM
2 Comments:
Part of the process should be figuring out if the decision can be automated not just if the data needed for the decision can be represented. Think about the decision you are trying to make. Can it be automated?
By
James Taylor, At
5:01 PM
James, excellent comment. . . I read your most recent post on Enterprise Applications and Decisioning (at http://edmblog.fairisaac.com/weblog/2006/07/enterprise_appl.html ) which gets to this issue. I'd drop your question in between steps 2 and 3, or possibly between 3 and 4 above, depending on who's making the determination on whether the decision can be automated.
Thanks for the comment!
By
bigskythinker, At
9:01 PM
Post a Comment
<$I18N$LinksToThisPost>:
Create a Link
<< Home