As an entrepreneur, you’ve likely had your fill of buzzwords. From bottleneck to Blue Sky Thinking, from Granularity to Red Flags, from SEO to Advertorial, your days are likely littered with such cliches. One of the greatest challenges is separating the useful, the actionable and the timeless from the fads and the deliberately bamboozling jargon. Big Data is one example of something that needs to be taken seriously… despite its somewhat obnoxious and buzzword-y title.
As a business leader, you take insight wherever you can get it. From books and movies to anecdotal conversations with your customers. You know that the right insight at the right time can help you to implement the policies and procedures that could lead to sustained business growth and prosperity. Especially in this difficult time when many businesses are floundering, all are trying to engage their customers in an age where consumer confidence is low, and all want to position themselves as trustworthy and earnest in the eyes of their clientele.
Big Data can be extremely useful in developing insights that can inform agility, growth and a superior service for your clients. But it can also become a colossal headache. Let’s take a look at some Big Data mistakes that could cost your business money and opportunities…
Having no structure or standardization in how data is collected
Collecting huge volumes of data is all well and good. But unless you’re able to mine that data for actionable insights, it’s literally wasting space in the cloud or your servers. The first step to making data usable is having a structured and standardized approach to how data is stored and managed. As this piece on bordereaux report data shows, insurance companies can miss out on extremely useful insights from claim data because their storage solutions are unstructured and non-standardized. The right Business Intelligence tools can add order and insight to your data.
Using data to confirm, rather than to discover
It’s nice to feel reassured. But all too many businesses use Big Data to confirm what they already know, using data samples to support their hypotheses. While this may be useful at investor meetings, it is a fundamental failure to use your data to its fullest. Data should educate, inform and even surprise as often as it confirms.
Focusing too much on departmental data
It’s understandable that departments will want to pay close attention to their own performance metrics and KPIs. But there is a danger in this of individual departments operating as almost independent data silos. They might know how to improve departmental performance. But they might have little or no idea of how their efforts are helping the business to improve its broader organizational goals and improve margin- the one metric to rule them all.
Limiting the data groups you pull from
Returning to the idea of using data for confirmation bias, when you’re pulling Big Data from small samples to support a hypothesis, you could find yourself in a state of willful self-deception, using misleading stats that could limit your perspective.
Relying on machine learning rather than human learning
Finally, data is a tool like any other. Machine learning can yield all sorts of useful insights, but above all, it needs to inform human learning, applied to the right problems at the right time by the right people.
Otherwise, what’s the point?
Please note: I reserve the right to delete comments that are offensive or off-topic.