When it comes to being predictive and improving enterprise performance, best-in-class companies are not focusing on any one method or technology specifically, but are taking multiple approaches, according to a recent report from Aberdeen Group, "Predictive Analytics: The BI Crystal Ball."
"A lot of companies are still in their infancy in terms of predictive analytics," says Aberdeen analyst David Hatch, who authored the report. "The best-in-class companies attempt to get a 360-degree view of their customers. So they integrate customer information from all systems -- not just customer service, not just the call center, and not just field notes."
Best-in-class firms -- which Aberdeen classifies as the top 20 percent of surveyed firms -- are advanced enough to rely on structured and unstructured data from external sources such as email and Web pages to get a total customer view. Since these external sources are Internet-based, they provide companies with the additional intelligence in near-real time. Hatch says that average companies and laggards don't -- or can't -- use external sources.
As the best-in-class companies develop a total customer view, Hatch says, they've achieved a return of $3.50 for every dollar invested in marketing in the last 12 months, compared to a return of $1.40 for average companies and a negative return of 90 cents for laggards. (Aberdeen's laggards make up the bottom 30 percent of the pool.) Best-in-class companies also have higher customer-retention rates -- 76 percent of them have a customer-retention rate of 90 percent or better during the last 12 months, compared to 12 percent for average companies and 4 percent for laggards, according to Hatch.
There are so many potential predictive KPIs, Hatch says, that each company will need to select those key performance indicators (KPIs) that are most beneficial for its circumstances. In order to gain on best-in-class companies, Hatch recommends that firms define KPIs, including:
- the number of customer incidents resolved in a given time;
- the number of adverse events discovered; and
- the number of out-of-stock items in a given time frame.
The next step, he says, is to establish a formal method for updating the selected KPIs based on changes to those circumstances. The ongoing review of KPIs will ensure they're suited to current customer demand. Additionally, Hatch says, the predictive measures must directly relate to the overall effect on the business performance. For example, in order to be useful, any measure of the "number of adverse events discovered" must connect to actions that will mitigate those adverse events.
Hatch notes, however, that predictive analytics has yet to achieve its full potential. Even best-in-class firms still tend to use predictive analytics merely to focus on the customer, despite the fact that the technology can improve other parts of their businesses. A heavy-machine manufacturer, for example, could use predictive analytics to get a better handle on machine failure (and warranty expense).
News relevant to the customer relationship management industry is posted several times a day on destinationCRM.com, in addition to the news section Insight that appears every month in the pages of CRM magazine. To contact the editors, please email editor@destinationCRM.com.