Some of us may remember a time when, if a business wanted to communicate with a supplier, an executive dictated the letter to a secretary. She (it was almost always a she) took her notes back to her desk, typed them up, and brought the piece of paper back to her boss to sign. He (it was almost always a he) might sign it immediately or let it sit on his desk for a few days before giving it back to her to put in the mail. At the other end of the lines of communication the same process would happen, dragging the process out for a week to ten days.
Many more of us will remember when the introduction of the fax machine seemed like business at the speed of light—until the Internet, email, and IM made that speed a practical reality. Near-instantaneous communications have helped bring business closer to the goal of being "real-time enterprises," a buzz phrase promulgated by consultants and analysts. But speedy communication is only one aspect of doing business in real time. Even more important is the ability for execs to find out what's going on under the hood in time to fix problems before they cause something to blow up.
Good News Travels
The downside all businesses grapple with is that more electronic information received sooner has actually led to a data explosion. "Business has become such a networked electronic phenomenon," says Forrester analyst Nate L. Root, "that the volumes of data created have already surpassed what humans can understand by themselves."
Most large companies already have analytics tools in place to provide business intelligence, but they tend to focus on only one aspect of the business. The finance department may use analytics tools that were built into its financial package, while the marketing department may use one off-the-shelf app for customer relationship management and another for analyzing response to its marketing campaigns. These applications certainly provide intelligence about different aspects of the business, but because they're connected to different data sources, they don't provide a complete picture of the entire operation.
The airlines' pricing departments, for example, use a variety of proprietary systems to try to understand how competitive their rates are. It wasn't until America West began using a low-fare search tool from ITA Software on its consumer Web site that it began to get insight into its range of flights and fares. "The travel industry is used to getting its content in pieces and parts," says Cara Kretz, ITA vice president. "Schedules, rules for fares, the availability of seats—all that disparate infor- mation needs to be put together in order to offer someone a fare. We've taken all that data and calculate and compute it, then display it in a visually clean way." That was good news for travelers handling their own bookings online; it was also good news for America West. It turns out, according to Kretz, that America West wasn't offering its Web users all the possible options for flying between two points.
The bad news is, ITA's search engine can't connect to airlines' traditional global electronic data interchange system. Neither America West nor many of the other domestic airlines have such powerful and easy-to-use tools to analyze their operations.
Talk To Me
If the airlines could turn that search tool loose to gather and analyze information from all their databases and ERP systems, they'd be a lot closer to true business intelligence. A category of software has latched onto the business intelligence moniker and aims to do just that. Business intelligence (BI) applications tie together data from across an entire company. Thus, they purport to enable the enterprise to make better decisions and to measure the results in a standard way. BI is the latest phase—and the latest buzz—in the ongoing development of ebusiness.
While companies have long had the ability to analyze past operations and performance, says Eric Berg, managing director of Technology Forecast publications for professional services firm PriceWaterhouseCoopers (PWC), "The notion of business intelligence has expanded beyond the focus on aggregated historical data." With BI apps, a manufacturer can track in real time how the supply chain is performing relative to what is anticipated. "If the model says a shipment of parts should arrive at the loading dock at 3 p.m.," Berg says, "you can have the system set up so that if it doesn't arrive by 3:30 p.m., someone gets paged so they can investigate and perhaps see what changes they need to make to accommodate for the fact that the shipment isn't there." Those changes could hit every department from marketing to sales to HR. "BI gives you contemporaneous information about your business, rather than the rear view mirror view," according to Berg.
According to Forrester, enterprise-wide BI applications, which it calls BI platforms, offer users:
• A unified data layer: BI platforms create
• A virtual data warehouse in which all corporate information can be queried, mined, and reported.
• Shared metrics: Departments work from shared assumptions and standard key metrics that are enterprise-wide.
• Centralized tools: Web-based interfaces that are shared across the enterprise foster collaboration instead of argument on whose numbers are correct, while freeing IT to focus on infrastructure, security, and account maintenance.
BI software comes in two basic flavors. You could call them ground-up or over-the-top. Ground-up solutions are built in along with the rest of the ebusiness infrastructure when a company installs a data warehouse. For example, users of Oracle's E-Business Suite can include the CRM Business Intelligence module as part of their infrastructure build-out. Over-the-top solutions, on the other hand, appear as a software layer on top of disparate systems and products from several vendors. These solutions pull together information from the various systems, clean and translate the data, then analyze the merged data as one piece.
"The biggest barrier to getting meaningful information is the fragmentation of that information across disparate systems, complex languages, and currencies," says Kurt Robson, vice president and chief applications architect for Oracle Business Intelligence Applications. But Robson admits that the Oracle solution is really targeted toward companies that are standardized on Oracle or that need to standardize and integrate a bunch of legacy systems anyway. Oracle pins its hopes on companies that have separate ERP, HR, financial, and manufacturing software systems choosing to scrap it all and start over with Oracle.
But that's not feasible for every company. For those not ready to build from scratch, over-the-top solutions can bridge the different systems. "To be successful, most enterprise applications will have to have a significant analytics component," says PWC's Berg. "We're seeing that more and more from the big packaged software vendors. But even if you standardize on one suite, you still have a mix of applications, so you will still see those specialty vendors playing a role in cross-application integration."