“Data enabled” and “data driven”. These are the latest buzzwords adopted by the purveyors of data analytics products and modern management. Conceptually, it sounds great. Every business has a ton of data floating around – data about operations, data from social media, data about products, and, especially, data about customers. If the business can just take all that data and somehow crunch it into a graph or a report then everyone in the company can make better decisions. Give the sales manager more data about their pipeline and prospects and they will generate more revenue. Let marketing see how people feel about a product on social media and they will be able to better engage potential customers and generate leads. If the CEO can get a dashboard of everything going on in the company from number of widgets produced to monthly expenses, they will run the company better.
Nice idea but for one small problem – data is not knowledge. It’s not even information. It’s just numbers. The number of people who like the company, the number of widgets that manufacturing can
make in an hour, or the number of dollars flowing into the company coffers. No matter how you present it, no matter how fancy the graph, no matter how much data you correlate together, the data is meaningless unless someone knows how to interpret and make sense of it in the context of the business. And, even when someone can make sense of the data, they still need to do something with it. One of the conceits of the big data world is that everyone can instantly know what to do when presented with the right data. That couldn’t be further from the truth. In fact, more data can confuse as much enlighten. What if you have conflicting data? Let’s say that the social media sentiment analysis tells marketing that people hate your product and company and want nothing more than to toss your CEO into a volcano. The financial data, on the other hand, shows record revenue and profits. What do you do with that? Ignore one or the other? See the sentiment data as a harbinger of a desolate future even though the present is pretty keen? Or is the sentiment analysis only capturing one small piece of the market that is meaningless to the business? It’s hard to tell when you just look at the data. It’s difficult to make sense of data unless there is a broader understanding of the data informed by qualitative information and experience. Even deciding what data to consider and what to leave for later is a difficult problem.
As an analyst, I see this first hand. We gather lots of data for clients and then help them to interpret it. Even so, there can be multiple interpretations of the same data. There are often indeterminate results that are hard to understand if taken at face value. What makes the difference is synthesis and context. By looking across many different data sets over time, we can get a clearer picture of what the current data really means. Seen through the lens of experience, organizational traits, and market conditions, the data makes much more sense. A graph or table full of data may say nothing or even be confusing without this lens.
There is nothing wrong with wanting data and analysis to be a key part of business decision making. Interpreting the data and placing it in context to make the best decision possible requires skill and experience. Before throwing more data at the business, first understand if the right skills and that expertise exists in the people who will have to use it.