Charting an Effective Big Data Strategy

Big data has proven to be a valuable business asset, but using it to gain competitive advantage requires the right combination of strategy, technology and execution. Many companies, in their excitement to use big data to solve problems in new ways, hastily search for the latest technologies promising to propel their insights to the next level. Once these technology components have been integrated, project leaders often realize that the organization’s expectations and outputs are misaligned. Alongside these disappointments, however, are a handful of success stories in which companies have managed to hit on the right approach to controlled data disruption.

As the life sciences industry continues to incorporate new capabilities like artificial intelligence and machine learning alongside big data, company leaders often find their efforts paralyzed by the many moving pieces. To name a few, there’s the sheer amount of data available, the thinking and analytics required to solve the right problems, and the challenges associated with mining existing data sources-not to mention settling on a technology and process that can best serve a variety of stakeholders. To overcome this paralysis, companies can start by considering how the outputs-the resulting insights-of the data will be used.

Many life sciences companies have implemented or are planning to implement a data lake. Now, imagine your enterprise data lake as a town under the jurisdiction of a town planner. The town planner needs to determine who will have access to different parts of the town, map out ways that the town can grow and expand, meet the needs of the town’s residents, and identify businesses that can be built to increase the town’s productivity. Similarly, if your organization is building an enterprise data lake, the blueprint should include who will have access to what data and insights, how the data and analytics processes can remain agile, how it can suit the needs of a variety of stakeholders, how it will be properly governed to preserve the data’s integrity, and how it will improve the organization’s business operations now and in the future.

Strategize first, shop for technology later

Just as the town planner wouldn’t start with a fully developed town map, your organization shouldn’t start by shopping for a new big data solution. Instead, establish a clear plan for how the technology will be used and who will use it before building and integrating a data solution.

Consider the success of Airbnb, a technology company known for its disruption in the hospitality market. The startup began with the end users’ needs in mind. It recognized an unmet need in lodging options for travelers, put a plan in motion and crafted a technological solution around that idea. An alternative to the traditional hotel model, the Airbnb home-sharing app is now in the hands of millions of international users.

In the life sciences industry, biopharmaceutical leader Amgen took a similar approach when it began its recent data infrastructure transformation. Unable to efficiently access, integrate and analyze large and complex sets of global data, employee production and company processes were sluggish and ineffective. The company abandoned the conventional technology-first approach to its data lake woes in favor of starting with the end users’ needs in mind and defining a carefully constructed master plan to accelerate the cycle time from drug discovery to commercialization. Amgen uncovered a thread of common needs across various departments and built a platform to easily access and analyze its commercial operations, R&D and patient data.

The Big Data success formula

Life sciences companies often struggle to identify and add the right data assets, access existing data, involve the right stakeholders across the organization, and partner effectively with the right vendors offering the right solutions. To build and sustain big data capabilities that contribute to the organization’s overall business goals, life sciences companies need a formula that calls for equal parts planning and execution paired with the right technology. But that combination will vary from company to company. Here are five elements to consider before undergoing a data transformation:

1. Define the strategy with your end users in mind. It’s important to determine the impact that the new technology will have on individual roles and plan accordingly. For example, to ensure that your company uses big data to solve the right problems in the right way, consult with the people and teams who work with the data every day to identify pain points and system limitations.

In Amgen’s case, the team turned to the scientists, business users and commercial teams to gather a “perfect world” vision for a new data solution, including a list of macro changes that would elevate the company’s command of data, speed up projects and enable data-driven insights. The business team identified its biggest challenges and then scanned all departments for overlapping issues before devising solutions.

2. Invest equally in sustainability and innovation. Data solutions need to be powerful and effective, but also agile enough to withstand future reconfigurations and changing organizational needs. By building solutions incrementally, companies can expand the data architecture and leverage new technologies when needed. For example, the industry is beginning to think about how an already established data landscape will handle a shift to an outcomes-based insight strategy that incorporates predictive analytics.

Of equal importance is balancing these investments in sustainable technology with a focus on bold and fast-paced innovation. Oftentimes, leaders stand in the way of step-change improvements as they wait for the perfect solution to come along. Companies need to make decisions quickly so that they’re continually experimenting and cutting their losses when they determine that a particular approach has no value.

3. Expand your organization’s capabilities with the right solution. By adopting the right technological horsepower, companies can achieve efficiency gains and fine-tune analytics. Highly educated scientists often spend significant time searching for public data, for example, when it’d be more productive and a better use of those employees’ skill sets if the company were to build a platform that automatically gathers public data. Establishing a single data access point and automating manual activities can slash the time needed to complete data-oriented tasks from days to hours, boost employee productivity and define new efficiency standards for the company.

Automation has solved many of the concerns associated with having humans handle data processes and allows companies to produce more credible insights at a far quicker pace. Each step of manual data entry and analysis activities requires verification, a process that is both time-consuming and tedious. Furthermore, ensuring that data has been handled and entered correctly can prevent costly mistakes.

Beyond improving core processes, new technological systems should be wired to assist companies in their access to-and ongoing pursuit of-more and more data. Companies are learning to access operational data currently not in use, leverage public data sources and tap into previously untappable data sources.

4. Establish effective governance to support your new solution. Big data technology should be open and experimental, allowing end users to tap into their own experiences and backgrounds to expand the impact on business operations. It’s equally critical to ensure that the data assets are properly governed: An organized and enforced data governance plan will guarantee that data is both reliable and useful while preparing the organization for future needs.

To be successful, revamping internal processes requires more than building the right solution. It requires a mindset shift. Companies will need a universal and holistic culture change, revamping who’s involved in the new data and analytics process, how they work, and how the resulting insights are disseminated and used. To ensure organization-wide trust and buy-in, a “concierge service” training model can be established to meet end user needs on demand. User champions can ensure that everyone is up to speed on new capabilities and can create momentum in the organization.

5. Maintain the near-term focus and nurture a long-term vision. Just as a town planner needs the foresight to secure funds to replace the main road’s sewer line in five years while recognizing the importance of attending today’s town meeting to debate the subdivision of a residential property, those overseeing the data lake need to maintain both a near-term and a long-term view of challenges and goals.

With technology advancing at a rapid pace, companies need to simultaneously consider evolving business needs while keeping their eyes on new ways to gather, store and unlock data. One way to do that is by maintaining a backlog of high-level business problems that can’t be solved in the present time frame but may be solvable with the right technology or strategy in the future.

During the last few years, life sciences companies’ ability to mine and act on progressive capabilities like big data has determined their competitive advantage. As internal data and other types of data begin to enter the picture, companies are bracing for new challenges and opportunities. The companies that effectively plan and execute on holistic, agile and end-user-focused data and analytics strategies will take the lead now and will find sustainable success.

About the Authors

Mahmood Majeed is a managing principal with global sales and marketing firm ZS and leads the firm’s global business technology solutions practice. Vickye Jain is a strategy and architecture manager in ZS’s Pune, India, office. Sandeep Varma is a principal in ZS’s Pune, India, office.