June 18, 2008, 10:38 AM — For a pharmaceutical company like Wyeth, no function is more important than research and development-the process of finding the new drugs that will lead to patents and profits. And for the information systems group that supports R&D, business process management (BPM) is emerging as a key technology and management strategy to make that function more efficient.
In fact, R&D's early success at using this technology and methodology to cut software development time in half has sparked interest from other divisions of the company that are looking to start their own BPM projects. But for Wyeth, the real payoff lies in BPM's potential to help the company define business processes and unify its information systems to break down barriers between organizational and geographic divisions and to improve collaboration and innovation.
"The demand is very high for connecting what used to be stovepiped systems," says CIO Jeffrey E. Keisling.
That demand is driven in part by the global nature of the pharmaceutical industry, in which virtual teams from different business units around the world work together to develop new drugs and related innovations. "As we develop products, that development is happening on a worldwide scale," says Keisling.
For example, one current BPM project targets the process for developing medication labeling documents, which involves collaboration across many stakeholders and approvals that have to be obtained from regulators worldwide, Keisling says.
The process is almost as involved as the application for a new drug approval, he says. Wyeth has to detail the composition of the medicine with molecular diagrams, explain restrictions on its use and document known drug interactions-all to produce the folded piece of paper you find inside each package. The application, which is still under development, will need to reach across R&D, clinical trials, and legal and regulatory review.
The fact that Wyeth would trust such a critical, regulated process to BPM is a vote of confidence in the approach, says Keisling. (A Wyeth BPM project was selected to receive a 2008 CIO 100 award.) So far, Wyeth seems to have avoided the technology and governance pitfalls that have dogged some other implementations (see "Business Process Management: A Hot Area That's Still Immature.")
Keisling says the main governance issue he sees is connecting the BPM expertise emerging within the company with the projects that will deliver the greatest return on investment. "Right now, our biggest challenge is portfolio management," he says. BPM is definitely a to-do list item for companies today. But many CIOs and businesses have struggled to make it work. Wyeth's experience offers a window on what makes a successful BPM initiative.
Put Business First
Part of the reason for Wyeth's success is that its BPM projects have been defined with the emphasis on the business process rather than on the technology. They are driven by a business mandate, either born out of a regulatory requirement or an internal need, Keisling says.
"Our business partners have absolutely no bias in terms of the tool we use, but they have a strong bias toward seeing that we deliver results," he says.
Wyeth's BPM initiative aims to fill in the gaps between systems, promoting smooth hand-offs from one to the next, and shifting more of the responsibility for defining processes to business analysts, rather than programmers. Although the emphasis is as much on the business processes as the technologies to enable them, products such as the Metastorm BPM Suite (which is used at Wyeth) help by offering a combination of visual process modeling, process modeling, workflow, automation and integration tools. The BPM software can orchestrate processes that cross multiple computer systems, taking advantage of Web services and other integration technologies to route transactions from one system to the next.
Another important goal of implementing BPM is to identify parts of a business process that aren't automated. Often, these are choke points where an employee is responsible for taking the information from one system, performing a manual task or analysis requiring human judgment, and then kicking off a process in another system. In these cases, the BPM tool itself can provide e-mail notifications and reminders, in combination with Web-based forms, to prompt workers to perform those tasks and keep things moving.
This layer of workflow automation also provides visibility into processes that otherwise would occur away from the watchful eyes of corporate information systems. Since Wyeth is in the highly regulated drug development business, having better documented processes and auditing of how they are carried out could help the company in its dealings with regulators.
BPM is also helping Wyeth improve the efficiency of routine administrative processes. For example, one of its BPM initiatives is related to research projects that Wyeth conducts with the help of physicians. Rather than dealing with regulated medical data, it is focused on improving the interactions between Wyeth Medical professionals and the clinicians they work with around the world. The BPM solution provides significantly improved levels of management, collaboration and timeliness of managing these clinical research studies. Previously, Wyeth R&D personnel used a variety of systems and tools to track the activities of clinical investigators, including the number of patients seen and whether their reports on those patients met the requirements of the research protocol. Clinical grant payments were also handled in multiple ways by the research teams at Wyeth, leading to a payment request in SAP. Using the BPM tool, Wyeth can now introduce business rules to initiate the workflow for seeking approval for payments.
"The value is in the process consistency. Rather than relying on individual knowledge to make sure things get done, we can rely on an automated, documented process to get things done," says Jazz Tobaccowalla, Wyeth's vice president of information services, the technology group that supports the R&D division. "This is particularly important as we're going more global with our workforce and trying to leverage every hour available in the clock. It gives us a consistent way of doing things and a way to capture knowledge-everything that we can lose when people leave or people forget."
Bust Silos and Tie Processes Together
For Wyeth, the decision to focus on BPM emerged from an analysis of where the research systems group was putting its energy. "The group I inherited had a big emphasis on software development, with the idea that we should build software from scratch where possible," says Tobaccowalla.
Tobaccowalla shifted the emphasis to buying and adapting commercially available software. Yet the classic "build versus buy" trade-off was only part of the story. He also came to the conclusion that there was too much emphasis on the transaction systems and too little on those that enabled the processes that were valuable to the company. "Development was focused on siloed, transaction systems. We had integration technology that moved data from one system to another, but we did not necessarily tie the processes across these systems together. What we really needed was for the process to flow with the data," Tobaccowalla says.
In the R&D division particularly, "there's certainly a greater emphasis on efficiency," Tobaccowalla says, "so consistency and process and clarity of who is doing what is important." For example, the BPM system can include process-monitoring rules that detect when a required approval is taking too long-perhaps because the responsible person is out sick-and notify another manager.
One of the first benefits Wyeth saw from R&D's BPM initiative was that software development time was cut in half. Tobaccowalla says that on the average project, the actual software development that would have required six months of traditional programming work can be accomplished in about three months with a BPM tool. (This does not include the up-front time spent defining how the process should work, which is sometimes the bigger part of a project.) So while he had only planned to tackle three BPM projects in 2007, he wound up with eight underway by the end of the year.
Even so, Tobaccowalla says he has not found that BPM completely eliminates the need for a software development effort on his projects. While a business analyst can do more of the up-front work of defining a business process, there is still "a little bit of classic IT effort" to integrate the systems that must work with the BPM software.
Wyeth has also relied on consultants with expertise at configuring the Metastorm software to produce the actual process models, which is still a little too much like programming for the average business user. However, Tobaccowalla says Wyeth is planning to purchase Metastorm's ProVision tool, which is designed to be a more business-user-friendly tool for visual process design and reengineering, with the ability to generate models that can be imported into the BPM environment.
"Maybe with that, some of the hand-offs will become easier, and some of the simpler automations we'll be able to do with the click of a button," he says.
Success Speeds Adoption
The R&D group's success with BPM has attracted attention from other parts of Wyeth. In fact, Tobaccowalla is in the process of establishing a BPM Center of Excellence (COE) as a way for his staff to advise their business peers how to use the technology effectively. That's significant because "we don't establish COEs very easily," he says. In other words, the company doesn't devote those resources to every new technology fad that comes along, only to things it believes are strategically important.
IDC analyst Maureen Fleming wrote a research report on Metastorm that included a case study on Wyeth (IDC is a sister company to CIO's publisher). She says some companies who adopt BPM start with a grand vision for gaining better control over all their business processes. Others, like Wyeth, start with a specific application that happens to be a good match for BPM. What can happen then, if all goes well, is that the approach goes "viral" and starts marketing itself.
"When you have a good experience with a deployment, and it's on time and on budget, the uptake is very good. The heads of other departments start looking at it and saying, 'I want one of those,'" Fleming says. "And I think that's what happened here, where the outcome was viral demand inside of Wyeth."
Tobaccowalla says that Wyeth originally hesitated over the decision of whether or not to license the Metastorm suite. An internal technology review committee questioned the need for the suite, given that the company already had several other products such as Documentum and SAP at its disposal with workflow capabilities. Ultimately, the project team was able to make the case that BPM went beyond traditional workflow to manage processes that have to span multiple systems and that the tools had enough potential applications to be worth adding to the company's existing technology portfolio.
Wyeth's BPM initiative beginning in R&D is surprising to Pramod Sachdeva, managing director of Princeton Blue, a systems integration firm that targets the pharmaceutical industry. "This is starting on the R&D side? I think that's tremendous," he says, explaining that R&D technology groups are often too focused on specialized informatics technologies to pay attention to BPM. And that's too bad. "There's so much value to be created on the R&D side in pharmaceutical companies" where BPM could make a difference, he says.
Potential customers tend to be skeptical of claims made for BPM that "sound too good to be true," Sachdeva says. But it's coming into its own as a technology and a discipline. What's changed is the quality of the tools, which simplifies the integration required to orchestrate processes across multiple systems. "It's only in the last two to three years that I've felt these products have reached the level where they can truly bring value to the business-and not just be another tool for IT," Sachdeva says.
Still, the tools can only do so much. Implementing BPM can be a way of identifying and addressing the gaps in a process that cuts across divisions. But those in charge of the different divisions still have to agree on how the new, automated process should work, Tobaccowalla says.
Lacking that, the result is likely to be "a layer of bureaucracy that nobody is interested in," he says. "So the real magic is getting the business process right."