I heard a marketer say recently, “Publishing numbers in a powerpoint deck does not make them true.”
We have never had greater access to data to make decisions. But it has also never been easier to cherry pick data to support whatever point we’re trying to make.
Being data-driven in our decision making increasingly means being data-skeptical.
Whether intentional or not, bias in how we work with data can skew results. There can be bias in the data we select and bias in how we interpret the data. And biased data can lead to poor decisions. This is particularly true when pitching an idea, where you’re motivated to convince an audience.
The rise in data-driven marketing is turning us all into DIY amateur data scientists. Many marketers weren’t trained in statistics and are trying to learn data science on the fly. We have to learn to recognize some of the more common biases in working with data and not take results at face value.
We have to become better skilled at pushing back on the numbers.
(Marketoonist Monday: I’m giving away a signed cartoon print. Just share an insightful comment to this week’s post by 5:00 PST on Monday. Thanks!)
Here are a couple other cartoons I’ve drawn on biases in working with data:
Ehud Shavit says
I believe the main problem is the mixed signal sent to less experienced people – we don’t like it at all when there is no data, but when there is data we find it hard to believe and hard to accept the interpretation. Truth is that we really want as true knowledge and deep understanding, but this much harder to explain then giving a checklist of needed data.
Chet Frame says
An engineering manager once drew a curve using a single data point. When I pointed out that any curve would have the same potential if drawn with the same data, he said, “…but I chose this one.” We asked for more data points.
Wayne Coleman says
Have h are similar things in academic setting of various types.
One person was promoted partly on class evaluations of only 3 people. When questioned, no answer but was passed through anyway. 3 monkeys come to mind.
Priyanka Jain says
Great post! Really Big data has a lot of capacity to profit organizations in any kind of industry, ubiquitously in world. it is a useful to decision-making and helpful to improve the financial position of any organization.
Priyanka jain – http://www.kyvosinsights.com
Jann Mirchandani says
Makes me think of Mark Twain’s “there are lies, damn lies, and statistics.”
You can really cherry pick anything. Trying to draw the RIGHT conclusions is tough.
Connie says
PLEASE tell me you are going to make posters of this one. I need this one loud and proud at my desk at work.
We often deal with graphs and PPTs because so often now data is drilled down for a very short presentation (as short as one minute, thanks Pecha-Kucha…). There’s no time for caveats, discussions of bias, or potential biases in the data. Sometimes even the goal of the data is skewed, and yet the execs need it to show a particular point. I find presentations to be the real reason so many flawed projects are started – the data is manipulated until it no longer resembles its original form, with execs trumpeting how awesome it is, and then the project fails miserably because of the manipulation used for justification.
Christopher says
After 6 long years in Analytics, what I can say is “Analytics” is the most abused word used in the world. Remember a saying which my grandmother used to say “Learn to walk before you learn how to fly”, but every one wants to fly and so comes the bias.