Predictive Analytics: Where Do I Start?

Considering implementing predictive analytics in your organization but don’t know how to begin?

Start with planning! I really can’t stress enough how important planning for the data is. Does your organization have the right data to answer the questions that you are asking?

Before you invest in a strategy or tools, ask yourself the following:

WHO…

• Will be doing the analysis?
• Will need to see/access the results?

WHAT…

• Data will be used?
• Problems will be addressed first?

WHERE…

• Will the data be processed?
• Will the data be stored?
• Does the data that will most commonly be used for modeling sit?

WHEN…

• Will we need these results in order to meaningfully improve the business process?

WHY…

• Do you think advanced analytics will benefit you? What is the expected value for the advanced analytics strategy?

Some other questions to ask are:

• Aside from tools, how are we going to facilitate getting data to the analytics group? Where will it be stored and how will the analytics team interact with it?
• How quickly do we need to be able to build and implement these models? What will be the expected payback so we know what a reasonable investment is?
• What tools do we need? Perhaps as part of this process you will be selecting an additional tool to enhance currently available functionality, but surely there are already tools in place for data analysis in your organization (perhaps the most popular one is Excel), but this is an important consideration in deciding what else to add to the toolbox. It is important to ensure that the new tools are complementary, not redundant, and preferably can easily share data.

I realize that I have really just posed questions and given no answers, but the reality is that the answers are different for every organization and these differences in answers mean that a completely different structure and suite of tools is appropriate. It is important to think about all aspects of building your analytics strategy at the same time and to not pursue one part without the others. For example, spending money to recruit and hire a data scientist when your organization doesn’t have easy access to the data needed for modeling or hasn’t committed to buying analytics software probably isn’t going to work out well. Similarly, purchasing predictive software without a plan for who is going to use it will probably also result in unused software.

So often I hear from clients who ask, “Where do I write the check for predictive software”, but this is not simply a software solution. If you don’t have an existing team of analysts, this is a growing opportunity and will be different from most other projects you implement from an IT side.

In the videos below, I provide some more insight into Predictive Analytics and how the practice can improve your organization’s ability to turn disparate, ever-expanding data into meaningful, trustworthy information.

You Don’t Have to Be a Data Scientist to Use Predictive Analytics

Using Predictive Models to Leverage Your Marketing Spend

Building Practical Predictive Analytics Solutions with SAP HANA

And be sure to check out our DFT’s Predictive Analytics Webinar Series. I provide more detailed information on what SAP Analytics tools are available, building analytics strategies, and how to design and implement predictive models. You can hear recordings from previous webinars, as well as register for the next webinar, “Implementing Predictive Models: Making Insights Actionable,” on Thursday, September 24, 2015 at 2:00 pm EST. In it, I will walk you through the two ways of implementing predictive models, including the benefits and drawbacks of each.

Click here to register!

 About Hillary Bliss
Hillary Bliss is a Senior ETL Consultant at Decision First Technologies, and specializes in data warehouse design, ETL development, statistical analysis, and predictive modeling. She works with clients and vendors to integrate business analysis and predictive modeling solutions into the organizational data warehouse and business intelligence environments based on their specific operational and strategic business needs. She has a master’s degree in statistics and an MBA from Georgia Tech.