The advantage predictive analytics offers over traditional BI is the way it presents information. BI has always been dependent on people who are willing to go through stacks of raw data in order to discover the information contained within.
This can be very tedious work and well beyond the interest level or skill sets of business decision-makers. As a result, BI often turns into just more B.S.
Predictive analytics tools are now presenting the information in a more accessible format: e.g., charts and graphs that make it easier for non-technical people to visually identify trends and statistics.
Business users are able to see the predicted outcome of various decisions, compare that with the costs to implement and determine which courses of action will produce the best ROI. And they’re able to do it in far less time than ever before.
Going back to the blue product example (see last month’s column), perhaps the total manufacturing cost of adding a blue SKU to the line is greater than the expected sales revenue. In that case, the user’s business knowledge dictates that adding a blue SKU, while desirable, is not feasible — especially if the comments also indicate that not having a blue product isn’t a barrier to continued sales.
The decision can then be based on facts, rather than gut feel or a knee jerk reaction to customer comments, and a potential catastrophe can be avoided.
With all that being said, organizations that are doing neither are faced with a choice: do they start with traditional BI and then ease their way into predictive analytics or do they make the jump to predictive analytics straight away?
More than anything, the answer lies in the company’s culture.
Predictive analytics requires a leap of faith that BI does not. The methodology is the reason. Remember that BI starts with your assumptions and then tells you if the statistical patterns are lining up with them.
Predictive analytics simply looks for patterns anywhere and everywhere and, as a result, may take you far a field of your normal business thought patterns. That’s a jump some organizations are willing to make, and others are not.
Here are a few questions to consider when deciding between the two:
How willing is my organization to break away from established ways of thinking? If the corporate culture is one of following the status quo because this is the way it’s always been done, predictive analytics will make your co-workers’ heads spin. Better to ease them in with BI first.
Do you have an analytic culture that uses a structured approach to business planning? The more analytical the organization, the better suited it is to predictive analytics because the people within it will know how to turn the data into useful information. When presented with the trends, they’ll know how to process them.
What is the established methodology for forecasting and budgeting? Predictive analytics could produce results that are significantly at odds with current forecasting techniques. If there will be great resistance to changing the way things have always been done it may not be the right option.
Do outliers (a few big sales, a few underperforming offices) have the potential
to significantly skew the big picture? Predictive analytics works best over a large group. If a few players can quickly change the landscape it becomes more difficult to accurately predict future trends versus historical expectations.
What about external factors such as business cycles, currency fluctuations and natural disasters? If that’s the case, you may need to have the resources to bring in weather data, government economic forecasts, etc. to provide the base of data necessary for the calculations.
Do you rely on early warning signals/indicators in your business to spot trends or identify potential trouble spots? If you do, moving to predictive analytics will be an easy transition since it quantifies these types of factors and helps you spot others you may not be considering.
Does your company have and follow a 12- to 24-month business plan? Surprisingly, many don’t. If it does and it requires future facts and data to develop, predictive analytics can help with that process.
Is the staff familiar with advanced statistical tools? Generally actuaries and economists are familiar with these types of tools, while operational and line managers are not. It’s definitely an easier sell if there’s already some confidence in advanced statistical tools.
Can those people become “super users” who mentor others into making the jump? If
predictive analytics is restricted to a very small group of users it will be difficult for the organization to see the kinds of results that are promised. It’s when predictive analytics permeates the organization and taps into its full brainpower that it really has an impact.
Can you identify three cost savings or revenue opportunities from your core business operations where predictive analytics will help your business? While predictive analytics can help you discover the unknown, it definitely helps to know there are opportunities already available.
Being able to point to areas where having that kind of information can make an immediate impact — “if only we knew then what we know now” — makes it easier to justify the effort and expense of adding a predictive analytics solution.
It’s possible with the vision of predictive analytics to put the telescope of history away and start seeing your business trends of yesterday, today and tomorrow at virtually the same time. There’s a great deal to be observed and learned, especially if you have eyes in the back and front of your head.
Mark Robinson is a Business Intelligence practice manager at Greenbrier & Russel, a business and technology services firm specializing in business intelligence, custom application development and enterprise solutions. He can be reached at [email protected].