Data reporting functionality has been used for many years within organizations as the primary mechanism for informing key employees (i.e. management and those with customer relationship roles) about the state of "now"-the status of current operational performance: customer service levels, campaign responses, and other measures. Separately, advanced analytics methodologies such as data mining have been providing enterprises with the capabilities to use historic information of what happened "then" to make predictions about likely future outcomes. The two worlds of capability are closely related but do not typically work together.
What needs to change? Reporting tools and predictive analytics are usually entirely separate products, probably from separate vendors and mostly likely utilized by separate groups within the organization. It's no surprise then that they do not work well together.
Let's look at the case of expiring inventory otherwise known as the Valentine's Day example. Roses sell at a premium on Valentine's Day. However, their value declines dramatically if they are not sold by the end of the day. In fact, if a florist has too many roses left at the end of the day, some of them are likely to become worthless as they start to wilt.
The tool the florist is most likely to use to decide whether to drop prices is a report showing the current inventory. The report may show large quantities on hand, and therefore prompt the florist to start selling the roses at a discount in order to minimize the risk of being stuck with extra inventory. However, if previous sales patterns were studied, the florist would know that there is a large spike in demand at the end of the workday due to the fact that many people procrastinate and pick up roses on their way home. This is exactly the time when the prices should be highest. A predictive model would reveal those patterns.
We hear from many of our larger customers that they know how to create predictive models that have value to their organizations, but they struggle with how to develop the models and deploy them more rapidly. At a minimum the output of the predictive models should be incorporated as part of the reports on current status. In the example above, the florist would have had a more complete picture and would not have panicked about current inventory levels. Conceptually, this is not very difficult in cases where the models and business patterns are relatively fixed. In the case of Valentine's Day the statistical specialists could spend an entire year developing and deploying a model based on prior years' sales patterns.
However, we know that we operate in increasingly dynamic and competitive markets. In the worlds of customer interactions-sales, service, and marketing-we must find ways to interpret the past and combine it with the current situation much more rapidly. Reporting and predictive analytics platforms need to be integrated into a single environment and need to be agile enough to quickly produce results that can be embedded directly into the business processes of your organization.
A second change that can help detect patterns and enable a more predictive enterprise is to enhance reporting environments with interactive visualization capabilities. The human mind is amazingly adept at spotting patterns-certainly in the case of simple patterns it can eliminate the need to create a "computerized" model to perform a prediction. We can look at the pattern of historical data and project what is likely to happen based on a current set of data. Sometimes these patterns are just beneath the surface or obscured by outliers in the data. Therefore with some additional interactivity, users can see slightly different views of the data that help them eliminate those outliers or identify different segments of the customer base quickly and easily.
"Now and then" some simple changes can have a big impact on the organization. Reporting tools and advanced analytics need to be tightly linked to enable a truly predictive enterprise-one that will improve customer satisfaction and enable better decision making.
About the Author
Dave Menninger has over 20 years of experience in the software industry and is currently vice president of marketing and product management at InforSense.
Please note that the Viewpoints listed in CRM magazine and appearing on destinationCRM.com represent the perspective of the authors, and not necessarily those of the magazine or its editors. If you would like to submit a Viewpoint for consideration on a topic related to customer relationship management, please email viewpoints@destinationCRM.com.