Demographics Makeover


Earlier I identified coloring the tips of the bars as a nice technique used by Quantcast to highlight over-indexing, and I took issue with their use of pie charts and the cryptic INDEX scoring system. In this post, we will examine those techniques and how they might be improved.

This view below of demographics data for the web property is the focus of today's makeover:

Pie Charts

There are many arguments in critique & defense of the pie chart, mostly in critique. Among them, Decision Viz has authored an entire series on "how to leave your pie chart":

10+ Ways to Leave Your Pie Chart | DecisionViz

Cole Nussbaumer is on a mission to rid the world of ineffective charts, "one exploding, 3d pie chart at a time".

storytelling with data: alternatives to pies

And Andy Kriebel at Facebook literally skewers people who use them.

Twitter / VizWizBI: This is what happens to people ...

Some will contend that pie charts with ~three evenly distributed categories are OK. I would argue, There is only one scenario where a pie chart enhances accurate cognition:

  1. Two categories
  2. One of them is by far the dominant of the two

For example:

Twitter / MonaChalabi: Amazing. Even if there are ...

People that door latches keep out

These visualizations communicate efficiently the dramatic difference between two slices of a single pie. Yet even here, accuracy and precision are lost without labels. In the latched door example, is it 3 % of thieves? Or four? Or 4.2%? Without a label you simply can't tell.

Best practice: avoid the use of pies charts. And on the rare occassion of a dramatic disparity between two categories, then a one-off pie chart with labels can communicate this very unique and specific message.

The Confusing INDEX

Cryptic as it may be, the INDEX system is "currency" in Advertising. Like many specialized industries, advertisers speak their own language. For the general public however, hiding numbers behind a derivative absconds the data.

Consider this example again: an INDEX value of 84 means that, "a given visitor to the website is 16% less likely to be of a certain demographic than any internet user chosen at random."

To most of us, this is confusing. 84 simply -16.

Admittedly, Quantcast is in a bit of a jam here. A portion of their consumers are in advertising. Those consumers demand the INDEX as a medium of exchange. And a healthy portion of their data consumers are laymen who do not speaka-da-language.

Drill in, and Quantcast do offer a view of the real numbers. But this too can be improved.

The Problems

The bars & pie chart above are redundant. They repeat the same information, both times inefficiently and also inaccurately.

In a bar chart, to visually represent the data, all bars must base from zero and their size must represent the accurate quantitative value. In the charts above, while each grey bar represents a differing numerical value, the US Average bars are all the same size.

The birthday candles distort the data-to-ink ratio and distract attention away from the data.

When used as a color legend (as done for the other demographic attributes like gender and ethnicity), the birthday candles are also inaccurate. They suggest that visitors aged 25 - 34 are the only "Adults" to visit, which simply isn't true.

And there is inconsistent use of color. The blue for "Adults" in the pie chart is used for over-indexing in the bars. So, across the vis, does blue mean "Adults"? Or does blue mean "over-indexing"? It cannot do both at the same time.

The answer = simplify

"Perfection is attained, not when there is nothing left to add, but when there is nothing left to remove."

-- Antoine de Saint-Exupery

The life of a designer is a life of fight against the ugliness

The single chart below conveys the correct information in a tightly compacted view. Lengths of the bars visually represent their quantitative values. For the laymen among us, the numbers are easy to understand. And the INDEX is available for the advertisers.

In Conclusion

By reducing total ink & consolidating two redundant charts into one *efficient* visualization, by removing the non-data clutter, and through a consistent & accurate use of color, the data themselves are revealed.

As a result, new real-estate becomes available which opens the possibility to present all of the demographics together in a single view, with a new focus on interactive features like filter actions, informative tool-tips, and comparisons.

And, of course, there is a time and a place for pie:

Word Count: 713


  1. "A Data Visualization Opportunity", Keith Helfrich, Red Headed Step Data, May 4, 2014:
  2. "Color the Tips!", Keith Helfrich, Red Headed Step Data, May 14, 2014:
  3. "Quantcast Data for", Quantcast:
  4. "10+ Ways to Leave Your Pie Chart", DecisionViz, June, 2014:
  5. "Alternatives to Pies", Cole Nussbaumer, Storytelling with Data, June 4. 2014:
  6. "This is what happens to people ...", VizWizBI, Twitter, May 20, 2014:
  7. "Amazing. Even if there are ...", MonaChalabi, Twitter, June 4, 2014:
  8. "People that door latches keep out", Truth Facts, June 12, 2014:
  9. ""Perfection is achieved, not when there is nothing left to add, but when there is nothing left to remove.", Antoine de Saint Exupéry:
  10. "The life of a designer is a life of fight against the ugliness", Catriona Cornett, Inspire UX, April 9, 2008: