Sometimes graphs are not more than pretty lines

Two weeks ago I published a blog post, “Moore’s law… and beyond,” about a presentation in which I used Moore’s Law, Metcalfe’s Law and the Chasm Theory to characterize the transition from a technology driven business to a value driven business.

With a bit of creative chartsmithing, I combined the graphs of these 3 famous industry laws into one, and by visually cheating with axes, scales, and representations I came to the observation that the chasm is actually the point where the transition from a technology driven business to a value driven business needs to take place.

Besides some positive comments on LinkedIn and a bunch of likes on Facebook ― my post was even republished by the World Economic Forum ―  I also received these critical remarks:

“Combining these graphs is actually ridiculous and leads to invalid deductions. Your conclusions are unsupportable because there is no data being presented.”

and

“Graphs without meaningful units, operationalized axes, and statistical analyses are just pretty lines.”

Both commentators were of course absolutely right. But the presentation I reported on in my article had never been meant to introduce a new, mathematically correct forecasting model. Its only aim was to deliver a message about creating market growth through user orientation. The medium is the message. And this medium worked very well for me. Remember that the title on top of this blog page is “B2B storytelling”. And that’s exactly what I tried to do when I joined the three graphs into one…

One more excellent reaction to the allegations above came from another LinkedIn reader:

“Letters are also just pretty lines, but the order in which they are put gives them context and turns them into words with meaning. Quit being so literal and enjoy the graph within the context of an accompanying article and as a way to illustrate something interesting.”

The online conversation about my blog post probably reveals a more fundamental discussion: when we create and deliver a presentation, should we stick to the hard facts and figures, or is a presenter allowed to “filter”, “frame” or “massage” certain data?

Here’s my opinion. First of all, you should never, never lie to your audience or present them with facts and figures when you definitely know they are incorrect. There is nothing wrong, however, with omitting unnecessary details, or framing the content to better align with the message you’re trying to convey.

A good example of this is the use of (financial or industry) analyst data in your slides. Although most of these analyst guys are to be considered trustworthy sources, in my blog post about “the incredible lightness of numbers” I illustrated that the figures they give may sometimes vary by a large factor. Of course, as a presenter, you’re allowed to quote only the sources that ― depending on what you want to show or prove ― mention the smallest of the largest numbers.

As a presenter you can also influence the audience’s perception of objective data. The case (by Garr Reynolds) I have outlined in my post “the duck and the rabbit” shows how a table may be (mis)used as an alternative to a bar chart to display hard numbers in a less dramatic or emotional way.

And, finally, sometimes you may assume that a theory or statement is true, until somebody proves you it’s not. Take the anecdote of the 17th-century Dutch painting “View of Scheveningen Sands,” created by Hendrick van Anthonissen.

whale_painting

Until recently, the whole world assumed that the people on the painting were actually staring at a deserted seascape… until the restoration of the artwork revealed a beached whale on the beach!

Living by numbers

In a business or technology talk, the audience is too often confronted with slideuments that misuse PowerPoint to present large tables with huge amounts of numerical or statistical data.

Numbers are indeed a powerful tool –sometimes even an essential one– to support your story and prove your statements. But, are you really sure that the folks in the room will understand and remember all the data you present? Numbers don’t resonate with people until they are placed into the right context and displayed in an appropriate format.

Here are a few tips for embedding small data and big data into your presentations:

  • Present only the essential. Select a few compelling facts or crispy figures that support your message. Don’t lose your time –and the audience’s– by reading the numbers out loud. Distribute the bulk data and the nitty-gritty details as a handout (there is a misunderstanding that the hard or soft copies you give away after your presentation always need to contain exactly the same information as the slides you projected on the screen.)

  • Use images and charts instead of digits. The human brain interprets every digit as a picture, so complex tables and calculations literally overload your brain. A good visual representation or an infographic is lighter to digest and will better stick with the public.
    There are zillions of alternatives to bar charts and pies to depict numerical data. If you need inspiration, have a look at this periodic table of visualization methods published by visual-literacy.org. And when you want to add animated charts to your PowerPoints, you might learn to handle data like Hans Rosling.

  • Illustrate your figures with metaphors. (Visual) metaphors help you to capture the audience’s attention and to convey complex concepts. They also let your listeners better grasp the order of magnitude and the importance of the (sometimes transcendental) numbers you may be showing them.

A final word of caution: get your numbers right! People won’t appreciate that you provide them with false facts. Also make sure you know the details behind the numbers and that you can explain or motivate the data you present.

More guidelines and best practices for writing stories about numbers and bringing statistics to life for non-statisticians can be found in the Making Data Meaningful series, published by the United Nations Economic Commission for Europe (UNECE).