How AstraZeneca Is Transforming IT

By Rob Wright, Chief Editor, Life Science Leader magazine

If you have any experience in technology, then you are probably familiar with Moore’s Law – the number of transistors on a chip will double approximately every two years. Interestingly, this exponential growth in computer processing power tends to decrease the cost of technology exponentially over time. Thus, the corollary for early adopters of technology is buyer’s remorse. For example, that iPhone 4S you purchased just one year ago for $199 now costs just $99.

Perhaps quick technological obsolescence is why many pharmaceutical companies have historically viewed IT departments simply as cost centers. Yet in today’s pharmaceutical world, IT is being viewed in a different light. Emerging markets and increased outsourcing trends are leading to the rapid spread of almost any size pharma company’s technological footprint. Increased communication, both internally and externally, is no longer just a common business requirement — it’s now a business differentiator, especially in the highly competitive drug development stages.

In 2011, AstraZeneca took a key step toward transforming its IT department when it convinced 20-year IT veteran Angela Yochem to leave her job at Dell and become the pharma company’s new CTO. But when you have a company as large as AstraZeneca, the question becomes, where do you start?

Tie Technology Spend To Desired Business Outcomes
Yochem notes that technology is increasingly being viewed as a key enabler of many different types of businesses, which is very different from how it was viewed in the past. “Historically, we spent a lot of money on technologies specific to a line of business, subline of business, or a small geographic region.” The result of this approach was duplication and redundancy, which meant it was not only expensive, but didn’t offer improved business agility. “It’s not good when you have umpteen number of different systems doing the same thing when you should probably just have one or two. If you have to make a change to a business process, having that type of inefficient system requires a tremendous amount of remediation work across many different system types,” she affirms. To avoid falling into these bad habits when determining technology spend, Yochem advises that you make sure you can trace every expenditure to a desired, known, and accepted business outcome. “There is very little value in a technology deployment that is not directly linked to an outcome.”

Creating The Right Team To Evaluate Technology Spending
Considering technology investment decision making requires a tremendous understanding of context and a broad perspective, one of Yochem’s first tasks in her new job was to assemble a 23-member team of senior leaders called the enterprise architecture board, or EAB. “It’s a terrible name,” she confides. “Because in many companies, the EAB is not a team of senior leaders, but a group of people who write standard documents that get put on shelves and are often never reviewed again.” That’s not the role of this EAB, though. She wanted it to be a group of people who would determine and prioritize technology spending for AstraZeneca. That meant EAB members had to be people who weren’t just business-savvy, but deep technology experts. They needed to understand not only the elements of a technology implementation, but also how to prioritize the project and the interdependencies between project components, which sometimes are not obvious. “Having strong technology expertise allows for accelerated decision making (e.g. knowing when to use commercially available technology versus when a custom solution might be necessary),” she explains.

The plan was to build a team of senior-level subject matter experts from each line of business, including procurement, portfolio managers, and even finance. “We built the EAB with such senior leaders so that significant decisions could be made without additional layers of approval,” she states. “In addition, when you have senior leaders who are personally responsible and accountable for deliverables, that becomes a great mechanism for creating buy-in across the board.”

When it came time to select team members, Yochem, being new to the company, sought recommendations from the CIO, the Information Services Leadership Team, and business line leaders. She would ask the person making the recommendation questions such as, “Tell me a little bit about this person’s history. Why do you want them on the EAB? What sort of history does this person have with this level of authority? Are they going to be able to make these sorts of decisions? Do you trust them to make these sorts of decisions?”

Simplify, Then Prioritize
Once the team was compiled, its first task was to identify what level of technology capabilities were required to enable the company’s business outcomes, which at the time, ranged between 5 and 10 for each line of business and functional area. As Yochem reviewed the number of business outcomes, she realized the importance of simplifying before prioritizing. She did this by first thinking about AstraZeneca’s information system (IS) capabilities as consisting of two categories — core and differentiated. “There are about 200 core capabilities,” she states. “To make them more easily adjustable and discussable, we grouped them into eight categories.”

• End-to-End Process Management
• Application Lifecycle Management
• Information Lifecycle Management
• Information Visibility & Exploitation
• Externalization
• Collaboration
• Consumerization
• On Demand

The process of grouping the core capabilities into eight categories provided the EAB with a good view of the desired business outcomes. “That’s when we could start the analysis process to see if we could link what technology capabilities, skills, processes, and so on would be needed to enable those desired business outcomes,” says Yochem. “That’s a big piece of work. In other companies I’ve seen it take a year or more, typically with a tremendous number of consultants involved.”

Lessons Learned The Hard Way
In creating the EAB, Yochem admits not everything went as planned. For instance, she quickly learned the importance of the team behind the team. “When I first put together the EAB, I failed to get the right number of people in supporting roles lined up. It is not reasonable to expect senior leaders to be doing a lot of compilation work and lower-level analysis work, after the high-level analysis is complete. For that, we relied on our chief architect, Mark Brogden, and his team that he was slowly assembling over the course of the year. We still got the work done, but it was due to some heroics on the part of Mark and his team working in the background,” she explains. “I think this put an unnecessary strain on that team.”

Yochem learned another lesson, one which she viewed as being even more significant. “I wish I had personally spent more time with each line-of-business head along the way and keeping them informed,” she confides. “Because while they were very receptive to the output (i.e. the multiyear plan showing what technology capabilities need to be in place to support the desired outcomes for each line of business) of the EAB and seemed very pleased with it, I feel the updates that I have given have been, in some cases, surprising to them.”

To avoid making the same mistake, Yochem recommends writing into your operating plan the time necessary to provide weekly or biweekly updates to each line-of-business owner on the specifics as to what is coming out of the EAB analysis. “I think richer and more frequent engagement would have made the process go more smoothly,” she says.

“When you have senior leaders who are personally responsible and accountable for deliverables, that becomes a great mechanism for creating buy-in across the board.” Angela Yochem, CTO, AstraZeneca

If you take on the task of changing how you approach technology spending within your company, Yochem has some concluding thoughts. Stay plugged into your team, paying close attention to how they are doing, so as not to take them for granted. “Very sophisticated, master-level enterprise architects are hard to come by,” she states. “It is something that you need to be very careful in sourcing. Invest legitimately and aggressively in that capability, because that’s where what appears to be magic to everyone else is, in reality, tremendously sophisticated analysis.” She explains that the analysis work required to identify the optimal end-to-end technology landscape over the next few years and identification and management of the various interdependencies across such a large estate can be quite tricky. The enterprise architects anticipate and measure impact, understand risks and trade-offs, and enable faster/better business decisions as a result.

The effects of AZ’s IT plan are evident when you look at how the company enabled significant differentiated solutions for its business during the rapid delivery of FIPNet (fully integrated pharmaceutical network) this year. “FIPNet allows our R&D staff to locate experts and potential collaborators, quickly connect with them in a seamless way, then operate on very large data sets together over the course of a multilateral collaboration — a particularly important model for our virtual iMeds,” explains Yochem.

For FIPNet, John Reynders, VP of R&D informatics, and his team were able to build or leverage core capabilities delivered for use across many lines of business, such as federated security, cloud-based email and document management, an AZ external connector (API), and interactive collaboration tools. Then they layered differentiated capabilities specific to R&D’s needs, such as those necessary to support real-world evidence collection and growth, real-time pattern identification and matching, and a sophisticated rules engine that allowed them to orchestrate events across the multiparty ecosystem. “Rolling out something like FIPNet without the target-state capabilities (i.e. the overarching capabilities identified as being necessary to deliver the desired business outcomes) would have led to an unnecessarily constrained and expensive point-specific solution, taking longer to deliver and difficult to extend or leverage for other collaboration models in and outside of our industry,” Yochem concludes.

CTO Reveals Useful Tools

For Angela Yochem, AstraZeneca CTO, the availability of some useful tools helped in her quest to streamline the company’s technology spending process. For example, frequently bringing together a globally dispersed enterprise architecture board (EAB) and its seven-member support team for in-person meetings can be costprohibitive. “We try to meet quarterly face to face and find the U.K. to be a fairly central location,” she states. But in addition to in-person meeting, the team also took advantage of virtual meetings, via a tool called Lync, a fully integrated conferencing, instant message, workspace communication solution. This allowed the EAB to set up quick, on-demand video conferencing from desks or phones, incorporating a variety of technologies, including whiteboards.

Another useful tool AstraZeneca put in place is called TrouxView, an enterprise architecture software that allows you to look across multiple assets and capabilities and manage the linkages and interdependency. AstraZeneca also employed a tool called Apptio, a real-time dashboarding technology that facilitates your ability to manage the cost, quality, and value of IT. “We paired Troux and Apptio, because it is important that we think about how we’re looking from the finance perspective, quarter-to-quarter,” says Yochem. “It also allows for some projection capability. The linkage of the two of them has been very useful in managing the overall project.”

The Sprint Approach To Analysis

Redesigning any process or department at a company the size of AstraZeneca is a huge undertaking. In terms of running, it would be a marathon, as opposed to a sprint. However, Mark Brogden, chief architect at AstraZeneca, introduced the concept of incorporating some sprinting within the marathon, specifically, when it comes to conducting analysis. According to Yochem, Brogden was charged with conducting the EAB analysis meetings. “At the start of the meetings, he would essentially say, we are going to meet for a day, and within this day meeting, we’re going to have five sprints,” she explains. He would then explain to the EAB how long each sprint would be for each topic. “By placing time constraints around the amount of time allocated to each topic and then cutting the discussion off immediately at the allotted time, every member of the EAB was incented to be as crisp and focused as possible,” says Yochem. Although, while cutting off the discussion at the allotted time is indeed important, agreeing that the topic will be resolved according to where the discussion lands at cut-off time is a strong motivator. In general, the pace at which the team was able to deliver a significant piece of analysis, pretty early on, was amazing.” Yochem advocates that if you are in the process of conducting team-oriented analysis, consider using a sprint-based approach to provide focus and make greater utilization of your most precious resource — time.