In this article I'd like to talk about predictive analytics, and more specifically, I'd like to discuss about why predictive analytics is not as widespread as it should be.

So first of all, what do I mean by the term predictive analytics?

Predictive analytics is more of a buzzword really. It doesn't refer to some kind of scientific field, but what it really means is using data science in order to predict things about the future.
For example, if you are interested in forecasting, if you're interested in predicting demand or sales around the product or service for the next month or weeks or whatever, then you could say that yes, you are interested in predictive analytics.
Or likewise, if you are interested in, for exp., predictive maintenance, that is, predicting whether a machine in a factory is going to breakdown before it breaks down, then this is another form of predictive analytics. Or in healthcare, predicting the risk of someone being diagnosed with a particular disease or disorder. Well, that's another example of predictive analytics.
And so some of those use cases, or probably all of them, they sound pretty exciting, right?
So we're talking about making predictions about the future.

So why isn't this type of technology more widespread?

Probably you hear it quite a lot, so you know it's out there. But from many conversations ahead, with managers, with CEOs, with executives, there are certain industries that are making heavy use of predictive analytics like finance and other industries that are still quite conservative about the potential of algorithms for prediction.

And why is that?

And my opinion is that this really comes down to the culture that different verticals have. So the culture in finance or retail might be very different to the culture in sports, for example. And this translates in different attitudes towards risk and towards technology. So in areas where the people in those areas are used to working with the latest technologies or they're used to working with mathematical algorithms, usually you find that they are more open to any kind of suggestion around to try out things around machine learning or AI, etc., whereas other industries are more driven by tradition.

Predictive analytics adoption

Surprisingly, there are many industries that have a lot of data, but you still find this sort of resistance.
So for exp., I've talked to many people in marketing and what I noticed is that while many marketers they're very open to using data science or even predictive analytics. In this case, many others they're not really interested and it's the same.
With some other industries like oil and gas or shipping for example or pretty much any industry where heavy equipment is an important component of the business can greatly benefit from predictive maintenance. However the adoption of those technologies still at its early stages.

Conclusion

So in the end, the point I'm trying to make is that we have technologies which can potentially transform industries. In this case we are talking about less downtime when we're talking about predictive maintenance or we're talking about more sales if we're talking about predictive analytics in retail.
In any case we're talking about greater efficiency and profits. But we see many industries shine away from that simply because they feel they're not ready yet.
And again, I believe that the only way to solve this is through education and as data scientists we need to be able to demonstrate the benefits of these technologies and make it easier for other parties to really understand what are the benefits involved as well as the risks. Because let's face it, data science, and not just data science, any kind of new technology that is being deployed might have some associated risks.
I would argue that data science is one of those types of technologies where in most cases there's not so much real risk other than the opportunity cost for an organization because an organization might have had decided to spend resources in some other way.
Obviously there are exceptions to this rule. If you're planning to deploy machine learning classification model for diagnosing diseases then obviously there's the risk of misclassifying someone.
But overall I'd say that most machine learning algorithms, most use cases in data science there are not really any big risks.
So in any way going back to the beginning of what I started to talk about, predictive analytics describes a particular type of application in data science and there isn't that we're not seeing predictive analytics everywhere as we should, we should be seeing everywhere because of its transformative potential. There isn't a sculptural change but I believe we're getting there, the industries are getting there, but also societies at large, they're getting more comfortable with the idea of prediction. Obviously you know how predictive analytics deployed at scale is going to affect society or the economy.
If you want to learn more about predictive analytics, make sure to visit this useful blog thedatascientist.com.
In this short session I just wanted to raise awareness of the topic and also the fact that many industries are experiencing resistance to change.
However, if you are interested in this topic and if you believe like I do that predictive analytics is eventually going to be adopted by all industries. Feel free to reach out to me to drop me a message. I'm very happy to have a conversation and answer any questions you might have.

Useful links in this topic from our community:

Top 5 Predictive Analytics Tools In 2021
What is the Future of Predictive Analytics in Finance
Use of Data in Insurance sector — Insurance Analytics