Nov 14 2006
Posted by Carlos Granier-Phelps and filed under Featured Articles, Ratings Analysis
About eight years ago, I was analyzing television ratings at a leading network in Venezuela… when I came up with the following graph (see below). Rendering took forever in Excel, but the resulting image was not only beautiful, but highly informative. I basically took the network’s audience share for every 15-minute interval throughout the day, for every day of the year, and produced an area map, where interesting patterns became obvious.
The colors you see indicate the audience share for each point in time… deep reds indicate less than 10-percent share, blues around 40-percent and greens above 70-percent. This market is mostly a two horse race, which is why you see such high percentages. From these graphs it was easy to see patterns (like the orange swath down the right side of the image, or the blue-green lagoons at the top left and the blue river down the right edge) indicating areas where the network’s programming was strong or weak.
With more computing power, I’d have taken this one step further, adding a third dimension to the graph to indicate the difference in audience share with our main competitor. That would have made a cool landscape… Soon we could even have the Himalayas indicating dominance across a time slot, the cold ocean depths (and, why not, a little model of the Titanic) indicating hopeless programming. In between, maybe some palm trees and a Corona.
Data visualization is a powerful tool, and when done correctly it can bring new insights into otherwise stale data sets.

Here’s a detailed view of the graph, showing the chart for ABC (income-level demographics), from 4:30 p.m. to midnight, from April 9, 1998 until April 30th of that same year. On the right is the legend, explaining the different colors.

I found these on an old backup DVD along with the original Excel files and thought it’d be nice to share them. Let me know if you like them. My brain seems to be wired for pattern-recognition, so I tend to enjoy and appreciate these graphs a great deal.
If you like multi-dimensional graphics, I highly recommend the work of Edward Tufte. Check out his website and browse through any of his books next time you visit your local bookstore.
Technorati Tags: Patterns, Data Analysis, Television, Demographics, Visualization, Digital Media Strategy, Edward Tufte, Ratings, Multidimensional Graphs, Ratings
4 Responses
alfredo
November 15th, 2006 at 8:04 pm
1You should sell those as art! Buenísimo.
researcher
November 15th, 2006 at 10:23 pm
2Hi,
I would like to contact you about the TV marketshare visualizations and data,
but I can’t find your email.
Please respond to my email -
researcher (Stanford University)
Carlos Granier-Phelps
November 15th, 2006 at 11:08 pm
3Researcher, I sent you an email to the address you left in the comment.
Meanwhile, I’ll set up a way for readers to contact me directly if needed. Thanks for pointing that out.
Amiyah Garley
June 14th, 2007 at 3:33 am
4This one makes sence “One’s first step in wisdom is to kuesstion everything - and one’s last is to come to terms with everything.”
RSS feed for comments on this post · TrackBack URI
Let me know what you thought about this article
Find interesting stuff I've posted
Search all my blogs
Subscribe to this blog
About me
RED66.com is a blog about Social Networking & Digital Media edited by Carlos Granier-Phelps. You can find more information about me here. If you have a project, idea, opportunity or link you'd like me to see, follow this link to contact me.
You can stay up-to-date with RED66.com:
via RSS or email,
follow me on Twitter,
critique my photos on Flickr,
connect to me on Linkedin or Facebook,
or check out my favorite links
on del.icio.us and ma.gnolia.
Browse articles by category
Browse articles by tag
You're visiting RED66.com. Special thanks go to: Design Disease for the original BloggingPro theme on which my blog's design is based, Alfr3do at Quieto.net for his savvy design ideas, and Mark James for his FamFamFam icons. All content is by Carlos Granier-Phelps (contact me) and some restrictions on its use may apply.