Posted by Milos Sugovic
It turns out Andrew Gelman, a professor of statistics at Columbia University, was right when crying out about the use, misuse and abuse of graphs in academic research. It seems to me like businesses aren’t much better. To illustrate, let’s look at this typical problem as a 6th grade math assignment:
Source: Opera Turbo White Paper
(a) the x-axis isn’t labeled
(b) the y-axis isn’t labeled
(c) there’s no unit of measurement
(d) one cannot deduce the difference between the red and blue line
(e) all of the above
The correct answer is “e,” all of the above. The scary answer is: I came across the chart in a white paper produced by a software company.
Last time I checked, white papers were the pieces of writing many of us turn to for rigorous business research, industry insight, and analytical leadership. Perhaps Adobe Acrobat is outdated on my PC and it’s just not rendering the graph correctly. (I’ve asked others to confirm the mystery that is this jumble of shapes and lines) So if that’s the case, please let me know.
Otherwise, I can’t help but question the credibility of this compelling “green zone,” and the research that supports its claim. Opera Software, in this white paper, claims “Opera Turbo gives operators the necessary tools to control costs and utilize network resources more efficiently.” To back this argument, it presents the attached exhibit, Figure 7, which shows the increasing difference between projected traffic growth and projected traffic growth optimized with Opera Turbo. This difference, they claim, is the “data savings” that a customer would enjoy after purchasing their software.
If the graph were fully labeled, it’d be a compelling argument. Heck, the green area could make up a data savings rate of 100% or more in a 30-day timeframe. That’s a lot of bang for your buck. Or their software results in data savings of 0.01% in 30 years. Which is it? I guess we’ll never know.
So what hides behind this piece of marketing material and the business intelligence used to conduct such projections? Of course, there's the carelessness argument, which would directly reflect upon their professionalism. Maybe it’s just an MS Word-Adobe Acrobat compatibility issue, in which case this software manufacturer should update its Acrobat. Or is the benefit of their software inconclusive, leading them to omit relevant information? I’m not going to call it either way; that’s left for you--the consumer--to decide.
But I will imply this: Market research is a difficult and challenging task, as data don’t always speak of the conclusions one wants to hear. And that stratifies, unfortunately, business intelligence into two groups: those with an imagination and those without. The second group genuinely embraces the power of market insight. If you truly belong in the latter, preserve your credibility and make sure you don’t pull any of the tricks of the former.
Comments