08/03/2017 | Written by: Jack Esselink
Categorized: Analytics | Watson
Share this post:
Analytics have been around for some time now in the business world, it all started when long lists of figures were produced by mainframes which were then analysed by number crunchers trying to make sense and some business value out of it all. Thanks to the rise of the desktop machines and spreadsheet programs everything became a lot easier. However, spreadsheets have great analytical capabilities, they also have some drawbacks, having one version of the truth is usually a nightmare. So… Enter the BI solutions like IBM’s Cognos, powerful, versatile analytics and reporting tools with capabilities to determine the right KPI’s and easily follow up on them with confidence that everyone is working on up to date information.
Great stuff… and it’s getting better, enter Predictive Analytics! Where traditional BI analytics give you great insight in what has happened and where you are at this moment, predictive analytics add a new dimension by giving insight in what is likely to happen in the future. In this area we see two totally different evolutions which both can bring tremendous value to your business. The first trend we would call the “data scientist” trend where very clever mathematicians and statisticians are using open source software like R and Python to find relevant information in big data environments. On the other hand, there is the “business analyst” trend, people with excellent business knowledge but less mathematical proficiency and interest want easy to use data mining solutions allowing them to harness the power of predictive analytics without having to write a single line of code. IBM SPSS Modeler is a platform where both worlds meet, data scientist can incorporate their R and Python code while business analyst can use standard algorithms in a fully supported way to build their own predictive models. Nowadays, companies embracing predictive analytics still have a competitive advantage but when these analytics become a commodity, not using them would be something like bringing a knife to a gunfight.
In the near past and still very true for a lot of companies, analytical systems were built by bringing together all the necessary components (databases, BI and reporting software, predictive analytics software, deployment elements etc.). Today however IBM brings solutions to the market which are purpose build for a line of business, e.g. Predictive Customer Intelligence (PCI) or Predictive Maintenance and Quality (PMQ) with add on capabilities which go even further and tackle specific problems within a line of business, e.g. PCI Next Best Action in banking. Most of these solutions are also available as SAAS solutions, taking away the hassle to invest in your own hardware.
And it doesn’t stop here, we are entering the cognitive era where thanks to Software As A Service (SAAS) solutions, IBM brings the power of supercomputers to everyone at a very low cost. A good example is IBM’s Watson Analytics which is a self-service BI solution with predictive capabilities which helps business users build analytics using natural language. If you want to try this yourselves, just go to Watson analytics there’s even a free version for you to evaluate the experience!
Traditionally, analytics were used mostly in finance (e.g. forecasting), sales & marketing (e.g. customer targeting) and risk analysis. Today analytics are everywhere, in HR, manufacturing (PMQ), customer service, security, crime prevention etc. And we’re far from the end of our journey… stay tuned.
Source: The original Dutch article has been published in Numrush Digital Newsletter