Finding a molecule with the potential to become a new drug is complicated. It’s time-consuming. And it can be frustrating. Fewer than 10 percent of molecules or compounds discovered are promising enough to enter the development pipeline.
F or more than 150 years, Pfizer has worked to identify compounds that advance wellness, prevention, treatments, and cures for disease. As a result, Pfizer has built a global portfolio of medicines, vaccines, and many of the world’s bestknown consumer healthcare products.
From Promise to Production
Getting from “promising” to product is difficult. And without the Pfizer business operations Resource management specialist team, headed by Michael Linhares, Ph.D., it would be even more challenging. For the past ten years, Linhares has been instrumental in Pfizer’s research information management evolution. He introduced the company’s Research Information Factory, implemented its global reference data project, pioneered the concept of data federation, and spearheaded other enterprise data integration projects - all in the name of making it faster and easier to obtain and use valuable data.
The pipeline from discovery to licensing occurs in phases over 15-20 years, and few compounds complete the journey. The initial study phase represents a multimillion-dollar investment decision. Each succeeding phase - proof-ofconcept study, dose range study, and large-scale population study - represents a magnitude-larger investment and risk than the one before. Senior management and portfolio managers need to know which projects the company should fund. Which compounds are meeting Pfizer’s high standards for efficacy and safety? What are scientists discovering in clinical trials? The data needed to support these decisions is critical.
Portfolio and project managers routinely make complex tactical decisions. From identifying projected costs and prioritizing multiple development scenarios, to assessing project impact on other team resources - having timely, accurate data is essential to making the best possible decisions.
Each study phase generates massive amounts of data and requires extensive analysis to provide an accurate picture. Data comes from Pfizer research scientists all over the world; from physicians; clinical trials; product owners and managers; marketing teams; and scores of different back-end systems. Over the past several years, Linhares and his team have transformed Pfizer’s ability to answer these challenging questions.
From Bigger to Better
In the past, research data resided in data warehouses and hundreds of Excel spreadsheets held by individuals around the world. Getting useful data to answer questions took weeks or months.
Bringing together the right data sources - even for a small dataset - could take a year and $1 million.
“You can’t run global research and development on spreadsheets,” says Linhares. “Valuable data was locked up in spreadsheets and transaction systems - inaccessible to people who needed it. We had to do something different.”
“Different” meant finding a way to deliver information in real time. It meant consistently slashing time-to-answers. It also meant greatly reducing the cost of finding those answers. Instead of building data marts with extraction, transformation, and loading (ETL) techniques that required a year and $1 million to deliver one answer, they needed a way to answer questions in days - or less - for much less.
Linhares and his team found their solution in Cisco Data Virtualization - agile data integration software that makes it easy to access and gather relevant data, no matter where data sources reside. The integrated data virtualization platform lets users query data across Pfizer’s global network as if the data resides in one place.
“Before, even when we were able to gather the right project data for decision-making, project teams often questioned its quality and reliability,” says Linhares. “Instead of being able to confidently move forward with decisions, we were stuck on debating where the data came from and if it was trustworthy.”
Now, research and portfolio data lives in one virtual place and provides “one version of the truth.” And it’s available to all of the teams that need it. Having fast good data allows Pfizer to recognize and prioritize opportunities faster.
“Sharing information in an open environment increased the value of our data,” says Linhares. “Of course we’re always working to improve data quality. But now, the data is readily accessible. It’s more reliable. And it’s much more actionable. Teams can immediately use it. Our portfolio and product mix has improved as a result.”
Having the data readily accessible, Linhares’ team now answers portfolio and business development questions much faster. Taking an agile development approach, with Cisco Data Virtualization they “triage’ an answer in an hour or two. They then develop, build, and integrate data sources in a week or less. Working with their internal customer, they then iterate the model and place it in production for large-scale collaboration within a month.
“Data virtualization also is far less expensive than building specialized data marts to answer questions,” says Linhares. “With Cisco Data Virtualization, our portfolio teams get answers in hours or days for about one-tenth the cost.”*
Data virtualization has definitely saved valuable R&D budget for Pfizer. But it also reduces risk and opportunity costs. Fast access to trusted data delivers deeper visibility into portfolio components. “We have adjusted our portfolio considerably based on data-driven decisions, and the savings are significant,” says Linhares. If one compound is not working out as hoped, Pfizer can quickly free resources to focus on another that is. Also, with fine-grained data readily available, high-cost decisions can be made more confidently.
“Data virtualization is far less expensive than building specialized data marts to answer questions. With Cisco Data Virtualization, our portfolio teams get answers in hours or days for about one-tenth the cost.”
Michael Linhares, Ph.D., Sr. Director, Business Operations Resource Management Specialists Group Leader
According to Linhares, with data virtualization, everyone owns the business. Data is shared, so his team’s internal customers can get the answers themselves. It’s easy to use. And it’s real-time. Business ownership has its privileges.
* Results are based on information supplied by Pfizer and may not reflect results attained in other institutions.
Customer Name: Pfizer
Number of Employees: 100,000+
Answering questions took months, slowing time to market for compounds
Creating data marts to answer specific questions cost millions of dollars
Valuable data was inaccessible and spread across numerous sources
Cisco Data Virtualization
Slashed time to find answers from quarters to hours
Reduced costs of data gathering analysis to one-tenth of former cost
Improved portfolio and product mix decisions
Today, information drives everything. The ability to use your business’ data assets to drive better business outcomes is what makes the difference between leaders, such as Pfizer, and second-tier or third-rate businesses. But disruptions such as big data and the cloud are tending to proliferate isolated data silos, slowing your business and IT teams’ progress.
Data virtualization technology hides the complexity of isolated silos. It quickly brings data sources together into a unified, business-friendly view of data, regardless of where it resides. Cisco wrote the book on data virtualization: Data Virtualization: Going Beyond Traditional Data Integration to Achieve Business Agility, and is the market leader with the largest customer base and most advanced users. Cisco Data Virtualization customers have been highly successful at achieving better business outcomes including:
Media and content delivery firm Comcast drove a $21 million in increased revenue from its installed base.**
Wireless technology innovator Qualcomm integrates data 10 times faster than using traditional methods.**
The NYSE Euronext financial exchange saved over $4.5M on its first project.**
** Data Virtualization:
Going Beyond Traditional Data Integration to Achieve Business Agility, JUDITH R. DAVIS AND ROBERT EVE, Feb 2014