Unfortunately, few companies have the luxury of replacing disparate technology with common systems across all of their units. Wendy's is a case in point. While all 1500 of the company-owned restaurants use the same technology, approximately 5000 franchises don't. The sales data that franchises send to corporate headquarters looks different from the data that company-owned stores submit because franchise data is reported on a weekly basis at an aggregate level. By contrast, more granular transactional data collected directly from the point-of-sale systems of company-owned stores is sent to corporate headquarters on a daily basis. As a result of those differences, Wendy's corporate doesn't have the highest possible level of visibility into its franchise operations.
Wendy's Deane acknowledges that this less-than-ideal environment for BI creates problems for the company when it needs to compare aggregated sales information from franchises with transactional data from company-owned stores - it's a hamburgers to cheeseburgers comparison. He says the company needs to increasingly make these comparisons as it looks to expand the pool of stores it uses for product testing and as it attempts to improve supply chain integration. To compensate for their suboptimal data collection environment, Deane is using an XML standard to collect more detailed information from franchisees who operate a large number of stores. (For smaller franchises, Wendy's uses a Web-based data collection system.) He also uses heuristics, or rules of thumb, based on activity at company-owned stores to extrapolate meaning from the aggregate data that franchises provide. For example, if a franchise-owned store does $US30,000 worth of business in a week, Wendy's corporate can make assumptions as to how that $U S30,000 would break down into sales of french fries, baked potatoes, hamburgers, chicken sandwiches and the like based on sales from company-owned stores in similar markets with similar aggregate sales histories. Proxies such as these may not be perfect, but they are a practical workaround and can be modified as needed to accommodate further integration with other systems, like the point of sale. Wendy's has no plans to get its franchises on standard technology because it sees its franchisees as entrepreneurs capable of making their own decisions about their operations, including choice of technology.
Because Wendy's is starting to understand the importance of having standard data to fuel business initiatives such as supply chain integration, the company was able to replace the phone lines and unstable modems that stores were using to transmit data to headquarters with a satellite connection in September 2002. The new, stable network helped improve the amount and quality of data that headquarters collects from both franchise- and company-owned stores. Where in the past Wendy's would miss information from as many as 40 stores out of 1200 due to unstable modems, it now gets consistent information from 1483 out of 1488 stores every night.
Why Force-Feeding Won't Work
Like so many technology projects, BI won't yield returns if users feel threatened by, or are sceptical of, the technology and refuse to use it as a result. And when it comes to something like BI, which, when implemented strategically ought to fundamentally change how companies operate and how people make decisions, CIOs need to be extra attentive to users' feelings.
When Wendy's began using its BI system to generate sales forecasts for stores, operators were sceptical. They didn't think technology could possibly take into consideration how local factors - such as weather, events and traffic patterns - affect their sales. Deane recognized that it's tough for people to quit relying on their experience and gut, so he listened to operators' concerns. Instead of forcing them to accept the forecasts, which he knew to be extremely accurate, he told them they could modify the forecasts from the BI system so long as they explained why and provided they later compared actual sales with what they forecasted and what the system predicted. The operators who modified the forecasts realized that the technology was often more accurate than they were. When they saw that they could improve their operations by better staffing their restaurants and more accurately ordering food to meet forecasted demand, they increasingly embraced BI. In effect, Deane let the users come to the trough on their own terms.
One might argue that Wendy's could have got better results more quickly had it forced store managers to use the forecasts. However, if it had, it would have run the risk of facing mutiny from the operators. And had store operators fought the forecasts, that would have disrupted operations much more than the delay the company experienced by letting operators modify the forecasts. Deane says being sensitive to users' concerns was more important, even at the expense of slowing down the rate of return.
"Trying to convince 1500 store managers to automatically accept a new tool that is going to have an impact on their ability to perform in their store is no trivial matter. You have to be very, very careful how you deal with the change management and the acceptance side of an implementation," says Deane. And if you do it right, you can realize an ROI of 430 percent over a five-year period, according to IDC. "Of all the projects that one attempts to do as a CIO, business intelligence, if well managed - and it's not always well managed - contributes far, far more than it costs," Dean adds.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.