The Truth is in the Data… Or is It?
"There are three kinds of lies: lies, damned lies, and statistics." --Benjamin Disraeli/Mark Twain
I often hear companies in various industries say that, “we’re a data-driven company” or “we live by our data.” This is a core principle acutely popular among technology companies – and taken at face value, this seems to make sense. After all, one would be remiss to ignore data crucial to the survival of the company – just as one would be foolish to ignore the data gathered from our senses telling us that avoiding the fire is crucial to our personal survival.
We live in an era of constant inundation from all kinds of data: big and little data, structured data and unstructured data, retail data, healthcare data, political data – you name it, someone’s collecting data on it. Even in sports, statistics (sports data) are ubiquitous. Twitter feuds are perpetually abuzz around game and player stats – with fervent recitation of data points used to bolster arguments that one player or team is better than the other.
What’s often missing in these “data” conversations is context. Context provides a more complete picture of the story – accounting for the circumstances under which the data has been collected that influence an outcome. A sales leader, for example, might compare the call volume of his sale reps from month to month and find a noticeable decrease coinciding with a decrease in sales. On the surface that leader might see this data and conclude that his team is slacking off and not closing deals. While this is certainly possible, it’s likely that additional context reveals other causes. Perhaps new legislation was enacted that impacted specific sectors or the overall economy; or the combination of regional inclement weather, holidays, unexpected budget shortfalls, and key people changing jobs led to delays any significant spending.
Some in the retail industry look at the continual struggles of some well-known, established store-front retailers and conclude that traditional retail is “dead”, and that e-commerce will put them all out of business. However, considering the successes of other traditional retailers offers additional context – revealing that e-commerce isn’t the only factor adversely impacting traditional retail sales; and that there are still many products people prefer to buy in a physical store versus online.
Context is what turns data into information. Data without context is meaningless, and at best, incomplete. It’s like taking photos of hundreds of individual faces and piecing them together to tell a story. The best you can do is say that it’s a crowd. Perhaps they’re excited or concerned; anxious or upset. Perhaps they’re gathered to watch their favorite team winning a game; or waiting in anticipation their CEO’s big announcement; or watching in disbelief their chosen politician losing an election. We don’t know. We don’t have the why, the when, the circumstances behind each face. All we have are the faces without context.
Despite this we often forge ahead making life-changing business and personal decisions and critical judgments based purely on data, not information – impacting business outcomes, careers, public and private policy, and lives. In Malcolm Gladwell’s The Tipping Point, he cites examples of decisions made based on data taken out of proper context – to the detriment of both a few and many – for what many perceive are “tipping points” for corporate, cultural or political change. Considering additional circumstances, or context, we often find another side to the story.
Ultimately, we often become beholden to data because it’s easy, measurable, and fits neatly into a spreadsheet. It’s something we can point to and say, “Aha!” It’s much more difficult however to gather the additional context – often a challenge to quantify -- to make a well-informed decision. Perhaps businesses should focus on being more of an “information-driven company” or a “contextual data-driven company” instead of a “data-driven company” – because the whole truth isn’t in the data; it’s in the information, the context. And while it’s not as compact or tailor-made for Excel, it does tell a more complete story.
Definitely we must look at context. Sometimes, in "data-driven" strategies we forget that context also provides critical data, and companies don't operate in a vacuum. I have seen the sales story played hundreds of times in real life, as if the sales person is all mighty being that can achieve any sales goal with pure will power.
You're definitely right in the importance of context. Often we make snap judgements without the complete picture. I reckon one more element that you're missing is the impact of user bias in interpreting data. We can be reading the exact same data set and recommend two completely different actions because of it. We have to realise our own limitations when being 'data-driven'.