Hang around business process management (BPM) practitioners awhile and, sooner or later, one of them will say, "It's a journey, not a destination." In a perfect world, you could design and automate processes and get everything right the first time. But in the real world, that's not how it happens. Even if nothing goes disastrously wrong, there's usually much room for improvement in a newly automated process.
The key to designing smarter processes is to measure and analyze process results to identify activities and steps that can be optimized. Business activity monitoring (BAM) and business intelligence (BI) tools used in combination with BPM software can give process designers the visibility they need to identify what they should be doing better.
Although the signs of convergence are just emerging at the product level, with integrations between BI and BPM companies and analytics and BAM embedded in BPM suites, more common are integrations and combinations of technology assembled by practitioners in the field. Where BAM is real-time and operational, helping you see what's going on in your processes to find and fix problems, business intelligence takes a deeper look at historical data and is more about optimization: What has the process done in the last day or week? Where are the hot spots? What processes are taking longer than others, or longer than they used to? What opportunities exist to make the process more efficient? You might spot a way to eliminate steps, or discover approvals are taking too long and talk to people about it or provide training.
Techniques and approaches often overlap, but here are five ways companies are using these technologies in combination to design smarter processes.
True to its name, BAM monitors processes, collecting key process metrics that can be visually presented in the form of a dashboard or analyzed in a spreadsheet. Many BAM modules automatically send alerts when a metric veers out of a predefined range.
BAM is so useful that some companies use it even before they get involved with BPM. "Let's face it, everything is a process," says Gartner analyst Bill Gassman. "Any time anything gets done, there's a process involved." Often, the first step toward better process management is monitoring certain stages or steps to determine how long they are taking, how often they're successful, whether they're meeting service levels and so on. Companies may then recognize the need for BPM software that will automate time-consuming manual tasks.
At the next stage, BAM can be extremely helpful in streamlining automated processes. "The combination of BPM and BAM gives business people knobs and levers of control over their business and awareness of opportunities or problems," Gassman says. Further evolution to business intelligence, he adds, helps business people understand where to position those knobs and levers for optimum performance.
Great Clips, operator of a chain of 2,500 franchised hair salons in the Midwest, was growing fast and struggling to share information across the many departments involved in opening new salons--legal, marketing, construction, management, operations. The company wanted to expand without hiring more people at headquarters. The first thought was that CRM (customer relationship management) software could solve the problem, but executives soon realized that CRM wouldn't take care of the human-to-human work handoffs or handle the sometimes complex relationships among various franchisees, many of whom had established relationships in different markets.
The company focused instead on BPM, custom database integration and document management. Using BPM software from Metastorm, Great Clips modeled and automated many of the 120 steps in the salon-opening process, and it now opens 250 locations per year, up from 150 previously. The firm was able to automate about 20 information-gathering steps and cut two weeks out of their usual 14-week new-store launch process.
Using BAM, Great Clips keeps an eye on automated processes, to find and implement improvements. "We never had good visibility into what was going on with the process," recalls CIO Jim Waldo. "Now we have good information on the status of work between key milestones and how those milestones compare to our standards."
For instance, Metastorm's monitoring module watches supplier delivery times. "One vendor was consistently late ... and that was leading to late openings in certain parts of our system," says Waldo. In other cases, Great Clips identified trends that the suppliers hadn't recognized, such as reorders for equipment that had repeatedly failed. "We use BAM to highlight those problems as they're developing and work with vendors to find the causes and rectify them," Waldo says.
A common way to use BPM and BI together is to analyze process or external system data using BI, to make more intelligent in-process decisions. This can be automatic. In a credit-card approval process, a service might pull credit risk information from a third-party source and, based on predetermined rules, determine whether the risk is too high. More commonly, a human being is kept in the mix, and BI is used to help participants make decisions.
TransUnion Settlement Solutions facilitates real estate transactions by giving lenders portal-based access to loan closing processes such as title searches, flood insurance and appraisals. The processes are heavily automated using a homegrown, J2EE-based system that uses Savvion BPM software for workflow capabilities. TransUnion also provides custom BI tools that help lenders select products and vendors, spotting which vendors are strong in which markets and in some cases taking the selection decisions out of the hands of a loan officer.
"In the traditional model, loan officers, often out in faraway branches, would each have their own network of appraisers and title shops," explains CTO Rick Carlberg.
As lenders attempt to reduce risk and ensure that their loan portfolios will stand the test of an audit for sale into the secondary market, they want to help their loan associates make more standardized decisions. Thus, TransUnion's BPM system can gather data about estimated property values, the size of loans and borrower credit worthiness and then apply a scoring mechanism to that data. Loan officers also can run the data through TransUnion's home-grown ePolicy BI tool and, depending on the lender's risk model, it will suggest which type of property evaluation product the loan officer should choose.
An automated valuation model (AVM), for example, is a quick, inexpensive evaluation that helps loans close in a matter of days. If ePolicy comes up with a low-confidence reading, the lender may upgrade to an alternative product such as a broker price opinion (BPO), in which a real estate professional visits a property to assess its value; this is still far less expensive than valuation by a licensed appraiser.
Every lender wants to drive down the time and cost of closing without increasing the risk of the loan portfolio, so Carlberg says that TransUnion "constantly works with customers to evaluate the outcome." Over a given set of loans, they'll look at how many of those loans were done by AVM vs. BPO vs. appraisals, and they'll make adjustments to the models accordingly. In this way, several lenders have significantly decreased the time, cost and head count associated with closing and managing loans.
One of the biggest obstacles to getting solid business intelligence is developing and analyzing consistent data. Shipping and logistics giant DHL International GmbH, which operates in 220 countries and ships 1.6 million packages per day in the U.S. alone, used BPM techniques to help develop the proverbial "single version of the truth."
DHL had a Siebel-based sales tool deployed all over the world, but the processes for gathering sales-related data and keeping data up to date weren't consistent or well understood. DHL practices BPM throughout the organization, and had a number of tools available, so it did extensive modeling to develop a new set of processes. Those processes are now followed by 8,000 salesmen in 105 countries and, because practices are consistent, the quality of the data is much higher.
"We now have probably the finest view of sales, but we have it the other way around as well," says Simon Bentley, vice president of the business IT strategy and planning group, in Scottsdale, Ariz. "We're now able to influence and drive our sales toward business goals far more effectively than we could before; we can up-sell and cross-sell with more efficiency and precision than we could when it was 8,000 separate Excel spreadsheets."
With the data-gathering processes aligned, BI analyses can give salespeople immediate feedback on incentives and targets, and also provide consistent visibility into global results. "Now we really are comparing apples to apples without having to go through some huge BI [data] translation," Bentley says.
Using BI in combination with IT management tools, such as HP OpenView, can help IT staff correlate infrastructure problems with process problems. For example, analysis of process data might spot bottlenecks that are tied to network problems or balky applications. If a one-minute process is now taking two, there may be insufficient computing resources or a disk may be full.
This type of integration is an emerging trend, and alliances are being forged between BI and IT management vendors. HP, for example, is working with Cognos, Business Objects and SAS to get these leading BI platforms to integrate with OpenView. Linking both tools with BPM systems and specific process data would complete the circle for IT-aware process improvement. Similar ties might also be forged with service level management software products from vendors such as Oblicore and Mercury Interactive. These systems look specifically at IT processes to identify ways to improve them.
There comes a time in the life of any process when one must step back and ask, "Is this generating the results we need?" Here again, BI software is useful, letting you analyze process metrics and related, big-picture data on cycle times, cost and so on.
At Great Clips, finding and deploying a BI solution is at the top of the company's agenda. The company needs to measure and analyze results that go beyond individual processes, such as profitability and complete cycle times, from concept to serving customers. Waldo says he wants to link process data with financial and point-of-sale reporting tools to create an overall dashboard. He's hoping for views at the salon, geography and franchisee levels.
"Looking at the overall ROI takes a much broader view than just the construction process," he says. "It takes all the activities of Great Clips into account, and to do that we'll have to deploy a heavy-duty BI approach."
Penny Crosman is the former Senior Technology Editor of IT Architect. Write her at [email protected].
It's easy enough to export process data to separate business intelligence tools. Or is it? Last year, a number of business process management and BI vendors formed partnerships.
Cognos, for instance, is working with EMC, FileNet, IBM and Savvion to link their BPM solutions with the Cognos 8 platform. TIBCO Software has integrated its business process suite with Business Objects XI. Metastorm and Hyperion Solutions are working on a similar connection, and Intalio is integrating its suite with Celequest's business activity monitoring software.
By integrating BI and BPM environments, you could launch diagnostic analyses in BI whenever process metrics fall out of tolerance. Conversely, events such as dashboard alerts originating in BI might trigger exception processes in the BPM system.
Several BPM vendors are developing or expanding BI capabilities built into BPM suites. For example, the just-released Appian Enterprise 5.0 BPM suite features "Analytics Everywhere" capabilities, and IDS Scheer's forthcoming Process Performance Manager adds BI features to the vendor's ARIS platform.
The market is moving in this direction, but the relationship between BI and BPM is too new--and out-of-the-box offerings too few--to call it a match made in heaven.