House Specialty: Fresh Data

Not all cheeseburgers are created equal. Just ask the management team of Red Robin International Inc., a privately owned, 170-restaurant chain that spent years trying unsuccessfully to come up with the right recipe for quickly assembling menus that appealed to consumers in a variety of geographic areas.

Red Robins problem was a common one: Sales, inventory, marketing and other operational data was housed in separate, aging databases. That made it difficult for the companys regional managers to get up-to-the-minute reports that told them which items were selling well—and which werent—in different areas. Sometimes, regional managers resorted to calling restaurant managers directly, then guessing at what would sell. Sometimes, they were right. Sometimes, they werent.

Finally, Rob Jakoby, the companys director of restaurant systems, and Howard Jenkins, the vice president of MIS, decided to start over, discarding the companys old reporting systems and cooking up a Web-based business intelligence system that consolidated critical decision-making data in a series of multidimensional cubes. The result: Red Robin has increased the number of high-margin items it sells while cutting costs by reducing food and supply orders.

In three years, the Greenwood Village, Colo., chain has seen its savings on food and supplies increase to $1 million per year.

For Red Robin, the transition to a new way of collecting and using BI wasnt smooth, however. First, IT officials had to convince the companys regional managers that the tools were an improvement over their ad hoc methods. And they had to clean up data in operational systems—a bigger job than anyone at the company expected.

Overcoming such obstacles, experts say, can pay off big, particularly for enterprises operating in increasingly competitive markets, where a few dollars saved or a few percentage points tacked onto the profit margin can make all the difference.

"The [BI] function is increasingly strategic right now because companies can do fact-based quantitative analysis rather than just going by their gut feelings," said Dan Vesser, an analyst with International Data Corp., based in Framingham, Mass. "It gives companies an opportunity to really look at a number of things theyve done in the past and realize that maybe theres a better way of doing something."

Vesser says the market for BI tools is booming. Worth $3.6 billion last year, it is expected to grow by at least 29 percent through 2005, IDC estimates.

Before and after

Before deploying the web-based BI system, Red Robins reporting setup was about as useful as a butter knife on a tough steak. Because the data often had to be pulled out of multiple, aging operational systems on demand, getting insights into business trends took way too long, Jakoby said. Not only that, but many regional managers had little faith that the information they received from the old reports was accurate.

"This information used to require a call in to the person who generated the reports, who then took 24 to 36 hours to get the job done, and then the data was out-of-date," Jakoby said. "Now, this information takes about a minute to retrieve, and its completely up-to-date."

Not only is it up-to-date, but its also one click away for regional managers, who can access it on the Web.

Red Robins BI system is based on the PowerPlay OLAP (online analytical processing) engine, Impromptu reporting tools and Upfront Web portal from Cognos Inc., of Ottawa.

Rather than taking the time to create ad hoc reports from a variety of operational databases, Jakoby and his team use the Cognos OLAP tools to bring together all the possible questions company officials will want answered, then map those to the answers in a multidimensional framework called a data cube. The key to the speed of this approach is that the questions are preasked and preanswered, so the system doesnt have to go hunting for the information. Users can access the information themselves using the Upfront Web portal software.

To test the idea, Jakoby hired Quest Database Consulting Inc., of Denver, to create a sample cube using SKUs and prices for food items. "We spent a token few thousand dollars upfront, but this demo really sold all the end users," Jakoby said. Soon after deciding to go ahead with the technology, Red Robin got Quest to train one of Jakobys staffers on the tools. In about a day and a half, Red Robin was able to begin building its own BI cubes.

Data cleanup

Before fully deploying the BI tools, however, Jakoby still had a data quality problem to deal with. It turned out that Red Robins regional managers had pretty good instincts because the companys data was, as Jakoby said, very dirty.

The company had a number of cross-referenced Microsoft Corp. Access databases that referred to the same products differently. An example: One food type, such as soda, was numbered 100 in one database and 200 in another, making it almost impossible to compare sales and cost figures. Another great (but scary) example: In the old days, reports on expenditures used to be off by a couple of thousand dollars each week, Jakoby said.

Jakoby and his team had little choice but to go into the operational systems and clean them up. "We sorted out our cross-referencing, and we threw a lot of data out," Jakoby said. Now, not only is it easier to create the data cubes, but also financials reports balance to the penny.

To date, Red Robin has created 10 cubes for everything from menu "engineering" to purchasing, food costs and financials.

The first cube was designed to help the restaurants regional managers better engineer menus by looking at their items as well as prices, quantity sold and the ideal cost.

Using this information, regional managers and others could click through the information and determine which items sold well and which had the highest margins. Based on that, the menus were re-engineered to reflect popular, high-margin items for various regions.

Another cube, which allows managers to track purchasing rebates, helped save the company $250,000, Jakoby said. Red Robin was running a special promotion on some of its drinks with a giveaway of beakers. Operations appeared to be running out, so the company placed an order for 250,000 more—at $1 each—only to then consult the cube, which showed the stores had mistakenly started the promotion a week early. Red Robin was able to cancel the order and save the money, Jakoby said.

The company is now building more BI cubes that will help managers track other areas of the business. Even a single percentage point reduction in certain costs can save the chain millions of dollars, Jakoby said. And thats a lot of cheeseburgers.