Operational BI: Digging Deeper for DataBy Bob Violino | Posted 08-27-2008
Operational BI: Digging Deeper for Data
Businesses rely heavily on data they gather from the field to determine whether they're meeting goals and customer expectations. Papa Gino'sis no exception.
The problem was gathering all that data caused some serious indigestion.
The Dedham, Mass., restaurant operator pulls in massive amounts of data--everything from statistics on how long it takes customers to receive pizza deliveries to how well restaurants stack up against the local competition.
Until May 2007, business managers culled together data, via e-mail, each day from sources like enterprise resource planning applications, restaurant point-of-sale systems and a slew of Excel spreadsheets. The process was difficult and time-consuming, taking hours as district managers, who are typically responsible for eight to 12 restaurants, accumulated data and passed it on to regional vice presidents for further analysis.
Then, Papa Gino's invested in business intelligence software from IBM Cognos. The application not only enabled managers to receive data much more quickly--it's now generally available to all key decision makers in the organization by 6:30 every morning--it virtually revolutionized how the company uses information to improve processes and better serve customers.
Papa Gino's deployed what's known as operational BI, which puts reporting and analytics applications into the hands of business users who can leverage information to work more efficiently and improve results. For many organizations, operational BI is taking data gathering and analysis to a new level of effectiveness. But it can also create new challenges.
Intelligence for Everyone
Intelligence for Everyone
Businesses decision makers need to have the latest and most relevant information to help their organization enhance processes and compete more effectively. So it's no surprise that demand for operational BI is rising.
In its operational BI benchmark survey, conducted in the fall of 2007, Ventana Research found that approximately two-thirds of the 314 organizations polled consider it "very important" to make BI technology accessible to all relevant functions in operations. Ventana found that 68 percent of the survey respondents have either deployed operational BI or have begun a deployment project.
The research shows that organizations want to increase the number of users working with BI software. "They want to push [BI] out to operational users who have been working primarily with spreadsheets or more primitive reporting and data access tools that are often part of departmental or vertical applications," says David Stodder, BI analyst at Ventana. Most organizations want to centralize the deployment of operational BI, so they can solve data integration, data quality and performance problems from a central hub, rather than through distributed and dissimilar applications.
Improving efficiency is a top expected benefit, according to 63 percent of the survey respondents. Organizations see improving the use and flow of information as an important aspect of extending the business advantages of using BI.
Also, organizations hope to be more efficient in how they deliver information to people. Ventana has found in its research that organizations waste time looking for information, rather than analyzing it. "Operational BI is intended to make it easier for users to access information," Stodder says.
Another key benefit of operational BI is that it can improve customer service. The users of BI in many organizations are customer-facing personnel, in areas such as call centers and sales. Operational BI should improve the quality of the information these workers use.
Organizations also hope to reduce costs with operational BI. They can do this through better use of information that will support managers' and users' insights into how to reduce costs in operations and business processes. They also want to reduce the cost and redundancy of having lots of disconnected silos of information, and multiple reporting and analysis tools.
Information to Go
Information to Go
Papa Gino's adopted operational BI as part of a strategic, five-year project to optimize IT systems and applications throughout the organization to improve the performance of its restaurants. BI became a high priority as executives realized the company could use the technology to more effectively leverage the massive amounts of data being gathered in the JD Edwards ERP applications; internally developed, in-store point-of-sale systems; and via spreadsheets. The business and many of the individual restaurants were performing well when Papa Gino's used the old process of data analysis. But executives realized that with the latest BI technology, they could not only improve the process and save time, but more effectively tap into a wealth of information to generate more improvements. "We wanted to turn the information into something more meaningful," Papa Gino's CIO Paul Valle says.
With the BI system, called IBM Cognos 8 BI, Papa Gino's managers use dashboards to quickly analyze financial data, such as revenue, at an individual restaurant by week, month or year, and how the revenue compares with performance at
the same restaurant in previous periods, revenue goals set by management, and revenue at other restaurants in the same region or state.
The system also provides reporting and analysis of operational data--such as how many customers visit a restaurant during various times, what sorts of items customers are ordering, and how many hours employees are logging--so managers can see how each restaurant is performing.
Food delivery accounts for about one-third of Papa Ginos' business, so a key statistic is on-time percentages for such deliveries. "We're looking at not just how often deliveries are on time, but what we're promising to the customer and how well we're hitting those promises," Valle says. So, for example, if a restaurant consistently promises customers 20-minute deliveries but delivers in a half-hour, that restaurant is earmarked for improvements.
Another major contributor to the business is phone-ahead orders, so other key statistics include how quickly order takers at restaurants answer the phone, how many calls are abandoned by customers, and how many callers receive busy signals when they dial the restaurants. "The industry standard is to have 85 percent of your calls answered within 12 seconds," Valle says. "That keeps guests happy and can create more orders by doing things more efficiently."
So if a restaurant isn't at least meeting that standard, Papa Ginos' managers can address the issue and make improvements. These types of enhancements can lead to increased customer satisfaction and more repeat business, the goal of any retailer. "We want to be best in class; to be doing what the guest wants from us," Valle says, and the insight provided by BI enables the company to do that. It also helps Papa Gino's restaurants operate more efficiently. For example, analysis of calling data lets managers know if there are enough people answering phones. When managers at the company evaluated data prior to the BI deployment, much of the business insight provided by BI wasn't possible. That means doing more than simply pulling in data--it's finding key metrics and making them available to business managers, via dashboards. In all, 80 to 100 managers at Papa Gino's use the BI application.
The company plans to move into the third phase of its BI implementation: using the technology to report and manage "by exception," Valle says. The first phase was to give business users the reporting tools and all the data they wanted to see. The second was to boil down information from dozens of metrics to 10 to 20 key metrics deemed vital to the business, giving business users the option to look at other data metrics as needed.
In the third and final phase, key metrics are examined for exceptional data. With this strategy, decision makers only look at data that's outside certain thresholds orpercentages--in positive and negative ways. For example, if a restaurant averages a total of daily guests that falls below a certain threshold, managers are alerted to the anomaly. In the same way, they're made aware of restaurants that have a higher than-expected number of guests. That enables district managers to quickly identify and nullify problems.
One of the tangible benefits of the BI system is that operations and finance managers spend more time analyzing data trends and less time collecting data, compared with the process used before the implementation.
Another benefit is that managers can use the forecasting capabilities of the BI application to get a better idea of how much product they should order and how many workers they should schedule, which improves the overall efficiency of operations.
As for performance improvements, Papa Gino's has seen gains such as improved on-time deliveries since deploying operational BI. "We're using it as a tool to help us refine and improve the restaurant experience, to increase guest satisfaction," Valle says. "If guests are not having a good experience, we're not going to have sales."
BI Plus GIS
BI Plus GIS
The Hillman Group in Cincinnati has used operational BI for more than three years and CIO Jim Honerkamp has seen several benefits. The company, which distributes hardware, signs, keys and other products, uses WebFocus BI software from Information Builders, providing applications to end users throughout the organization via an internal Web portal.
Hillman customized the BI applications on what end-users are trying to accomplish. One of the goals of the implementation was to shrink the time it takes for business decision makers to get information they need. Business managers in the past asked IT to run queries, and often multiple times in various scenarios. The process could take weeks. Now, Honerkamp says, they can get it within minutes. Hundreds of Hillman employees have access to BI, including people in finance, materials and executive management. Nearly 600 sales representatives use BI to track sales transactions and see if they're meeting quotas.
And BI helped Hillman avoid a costly fine from one of its customers. The customer's data showed that Hillman had not shipped product within certain parameters, making the company vulnerable to a $130,000 fine. But using data from the BI system, Hillman showed that the customer's information wasn't correct, and that Hillman had delivered the product within the set parameters.
Recently, Hillman added a geographic information system capability to its reports. A BI report might include a map of major bodies of water in the U.S., and a user could drill down to a specific geographical area and see information on all Hillman retail customers within a five-mile distance of a particular body of water, including how much of the company's stainless steel products have been sold in particular stores.
The company operates 10 distribution centers in the United States, from which it ships products to retailers like Lowe's and Home Depot. Using BI and GIS, Hillman managers determine the most efficient shipping routes to reduce freight costs. By optimizing shipments, the company has slashed total shipping expenses by 25 percent.
Hillman plans to move to more predictive use of BI so it can track key indicators such as the price of raw materials from suppliers overseas, and price its products accordingly in order to maintain profit margins.
Also leveraging operational BI to get more value out of information is Columbia Valley Community Health, a Wenatchee, Wash., health care provider that began using a product called Edge from Business Objects in 2007. The key driver was the desire to leverage clinical data to better manage patient care, says CIO Mike Hodgson.
For the most part, the BI system is being used by Columbia Valley's IT staff to support internal users. But the organization is preparing to roll out the application to users outside IT, including people responsible for monitoring information related to health care quality, such as patient outcome data and immunization programs.
Some of the most tangible benefits have been found from pushing data to non-technical users. For example, diabetes case managers receive a daily report alerting them about diabetic patients who are scheduled for visits that day, giving them a status report on eye and foot exams and the most recent lab results. "The case managers can then work with individual medical and dental providers to ensure that the diabetic patients in clinic that day can get up to date on labs, exams and education," Hodgson says.
Organizations looking to implement and use operational BI can expect to face some hurdles. For one thing, they need to provide strong and flexible reporting. Companies dissatisfied with their initial deployments of operational BI say having to redo reporting based on misunderstanding of requirements is a key problem. Deployment delays, difficulties integrating multiple data sources and the inability to get complete views of key information are other drivers of dissatisfaction, he adds.
CIOs need to identify the causes of delays and address them, and develop a strategy for creating an information management platform that can integrate data sources.
An issue that often comes up is data quality. While this is largely related to database maintenance, it ties in to BI because these applications draw data from these resources. Data cleansing--ensuring that the data is free of inconsistencies-- is something organizations need to be diligent about.
Hillman has created a data management organization that's responsible for overseeing the integrity of the company's data, including data cleansing. "There are 70,000 [stock-keeping units] we manage on three legacy transaction systems, and making sure the data between those systems is in sync was a big challenge," Honerkamp says. But data cleansing is essential to getting the most out of BI, he says.
Papa Gino's made sure to have the IT department clean up data in its various information sources prior to using BI. Before joining the company, Valle had worked as a consultant and helped several clients implement BI. He says by far the biggest delays in implementing the technology stemmed from companies thinking they had clean data when in fact they did not.
Given the potential complexities of deploying BI throughout an organization, Valle recommends that CIOs bring in experts to help with the project. Papa Gino's hired a BI consulting firm to help with implementation. "We knew it was a sizable project, with all the programming and data points," Valle says. "It would have taken us two to three times longer [to complete] had we done it ourselves."
The biggest challenge for Hillman Group, ironically, has been getting business decision makers to take full advantage of the capabilities of BI and act on the information they're receiving."We continue to offer up new possibilities [through BI] and we're getting a little bit of push back from some executives," Honerkamp says. "They say there's so much [information] out there today they don't need more. They're not quite sure which information they should act on."
Another issue is potential overload of the BI applications. Too many frivolous queries can slow down systems, Honerkamp says. The company uses a maintenance program from Information Builders that enables Hillman to monitor the use of BI queries and alert managers when certain thresholds are reached, he says.
Despite these hurdles, organizations are finding that operational BI gives them more valuable insight into operations and performance than they've ever had.