A Real-Time Timeout?

The use of real-time analytics has long been seen as a holy grail for data-driven applications, including CRM. But according to a new study by market research group Evans Data, although the number of companies using at least some real-time data has increased over the past year, many are learning that not all decisions need real-time data. The "Database Development Survey, Summer 2004" makes a distinction between real-time data, which is updated within the hour, and near-real-time data, which is updated within 24 hours. Using those definitions, 93 percent of companies said they provided real-time or near-real-time access to at least some of their analytics-capable data. Last year, 88 percent of companies said that was the case. Joe McKendrick, an Evans Data analyst, says he finds it "remarkable that so many companies report that they're as far along as they are in terms of at least having the data available for real-time analytic functions." The most significant aspect of the findings, according to McKendrick, is that companies seem to be taking a more holistic view of the information contained in their data. "There's a realization that you can't look at just one segment of data and take action without looking at other aspects of the business," McKendrick says. Citing just one example, he says, "if you're tracking sales data you need to also look at the number of returns at the back end...to really get a sense of how profitable a product might be." Much of the early application for real-time data has been to manage the supply chain or technology infrastructure, to provide alerts when bottlenecks begin to form. Today's real-time and near-real-time efforts, McKendrick says, focus on "what's now commonplace for system administration and moving that same model up to business administration." Unfortunately, as McKendrick notes, that kind of expansion comes with a hefty price tag. "It costs a lot to have a system like this in place, " he says. "The closer you get to real time, the more expensive the hardware, the software, the people to build and maintain it." As a result, he says, "most companies now are satisfied with near-real time. If you get close to the split-second nirvana, it gets really expensive." But some areas of CRM are more dependent on timeliness than others. "The contact center is a real good place to have as close to real-time capabilities as possible, especially in cases where you have customers calling in to check on the status of an order, or in the financial services industry where a customer might call to check on a transaction," McKendrick says. "The person they're interacting with needs to have up-to-the-minute data to help the customer with their issue or do an upsell or an upgrade." McKendrick says he sees a leveling off in terms of demand. "We found that the frequency of updating analytics-capable data within an hour has hovered around the 30 percent mark," he says, essentially splitting the difference between the 32 percent of respondents last year who said they updated data in real time, and the 29 percent who gave that response this year. Some of the disparities are industry specific. In e-commerce, for example, no respondents claimed to be analyzing real-time data, and only about 10 percent of retailers said they had any real-time capability. By contrast, more than half the companies in the information and medical industries claimed to be using real time at some level. But it's important to remember that "the market's still very much in flux," McKendrick says. Roughly 75 percent of companies are updating their data at least once or twice a week, a figure he expects to rise, he says. "A lot of companies see the relevance of using real-time analytics in some segments of their business, but it's not widespread by any means," he says. "The whole idea of delivery of real-time data is very appealing...but once they see the price involved, they change their minds....The closer you get to any kind of subsecond real-time delivery, the curve-of-cost just skyrockets up." Related articles: Does Access to Real-Time Data Make a Difference in Customer Profitability? Is Real-Time for Real? CRM Drives the Real-Time Enterprise: What Lies Ahead?