A Prediction about Predictive Analytics for 2016

2016

Since the start of the new year all I’ve heard are predictions. It may have something to do with it being an election year in the US, as well as an Olympic year. Elections and Gold Medals aside, if you’re reading this blog post, the most important thing about this year for you may indeed involve a prediction.

So, I will also join the prediction game. I am predicting that 2016 will be the year when Business Intelligence and Predictive Analytics merge forces like an inner join. But haven’t they done so already? No, not really. So why is now any different? It’s actually quite simple. For the last couple of years, I’m willing to bet the following conversation took place on a project you’ve been working on:

“Hey, wouldn’t it be great if we could forecast what our costs would be for the next three months and include them in our Dashboards? That would be very cool. Too bad we don’t have a data scientist in our department to do that for us!”

Likewise, if you were a statistician or a data scientist I’m pretty sure you’ve said the following to yourself:

“Hey self, you know what would be really cool? If we could actually have a program that automatically mines through data and cleanses it for us each month. And you know what would be even cooler? If we could just tap into it and run our forecast models without having to manually export and import large spreadsheet! Yeah self, that would be awesome!”

datascientist2

According to a New York Times article in 2014, a data scientist spends anywhere from 50-80% of their time ‘data munging’ through spreadsheets before any significant exploration is ever performed. It turns out that one of the main problems data scientists face on a daily basis is dealing with ‘dirty’ data and removing outliers and abnormalities. This is a specific business intelligence skill set commonly known as ETL (Extract, Transform, Load).

Coincidentally, at the same time BI folks are being asked to predict outcomes from the data they are producing on a regular basis for months and years to come and include them in their existing dashboards and reports. While not every data scientist is required to learn ETL nor does every BI developer need to  build a regression or forecast model, the skills needed for both are merging into single projects and will eventually merge into a single resource. That is why today, in 2016, if analytic companies are not pushing forward predictive functionalities onto their existing products, they are falling exponentially behind.

1b53d51dcb1bae1f6a351d4a881a5b71.548x640x1

The truth of the matter is that both fields have needed each other for some time, yet the technology has only recently caught up with that need. Over the course of 2016, EVTechnologies will be showcasing the many ways you can incorporate predictive analytics inside of your current Business Intelligence landscape. This can be done for companies that have SAP HANA or those that don’t have any Big Data solutions. Additionally, these solutions are not just for the data nerds but also for the sales and marketing folks, too!

Stay tuned!

References and Interesting Links

For Big-Data Scientists, ‘Janitor Work’ Is Key Hurdle to Insights

Learn more about SAP Predictive Thursdays

SAP Predictive Analytics Product Tutorials

 

 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.