Predictive / Data Analytics Gold Rush
Over the last few years, there has been a rapid growth in the use of predictive analytics to improve decision-making within companies. There is no doubt that predictive analytics will radically change the way companies operate and achieve goals. That said, I believe that we are embarking upon an “Analytics Gold Rush” in which many participants will engage in analytics projects with little to show that drives improved business outcomes.
I have been observing three trends which are leading me this view. These trends are listed below:
- The dawn of the Data Scientist aka Data Guru
- High rate of technology change
- The growing view that models can predict anything, even black swans events.
The dawn of the Data Scientist aka Data Guru
If you are embarking on an Analytics effort without analytical expertise on your team, it would seem like a fool’s errand. That said, I find it very surprising the number of “data scientists” and ‘predictive analytics experts” that have surfaced over the last few years. In the book “Outliers’, Malcolm Gladwell asserts that it takes about 10,000 hours of practice to become an expert. Although the 10,000-hour number is somewhat disputed, I think that that most would agree that mastery of complex skills isn’t quick. So where have these “experts” come from in such a short period?
The reality is that the individuals/teams that are needed must have the industry domain expertise coupled with the analytical background to be truly successful. However, what exists now is a big hype cycle around Analytics which has led to many vendors professing “deep expertise” to drive business value. You can see this hype cycle in a few different ways
- LinkedIn’s Data Scientist Group has 8800 members
- The hashtag #datascience is continuously used
- Every major consulting house has an analytics practice
- How many people that you know aren’t doing/thinking about some analytics / big data project?
All this means is that market participants will need to pay a premium for talent to pursue analytics projects which by their nature are risky. The risk stems from the likelihood of not finding any new insights or non-actionable insights from analytics projects. The following cartoon from Tom Fishburne provides some perspective.
Many will say that the risk/reward premium is worth it given the upsell potential, I have no facts/data points to dispute this, but the cost of failure will be a heavy one to bear.
So, given the complex nature of the work, leaders need to focus on assembling the right teams with the right mix to skills. In my view, the data scientist only one of the key members of a successful analytics team which is made up of individuals of broad and diverse in backgrounds and skills.
Over the next few posts, I will provide some additional thoughts on the rate of technology change and the desire to predictive everything.
The future clearly belongs to algorithms.
As always, my opinions are mine and don’t reflect the views of my employer.
A parting thought …
“I think data scientist is a sexed up term for a statistician … Statistics is a branch of science. A data scientist is slightly redundant in some way, and people shouldn’t berate the term statistician.”
Nate Silver
