Blogs

This blog is adapted from Building A Data-Driven Enterprise Through Strategy, Culture, and DataOps (ebook) recently produced in conjunction with Forbes Insights.

To many, becoming “data driven” is just another challenge on a long list of challenges being faced by organizations as they try to adapt to a rapidly changing, information-rich environment that affects virtually all aspects of the modern business. The flood of data cannot be escaped and is difficult to manage – yet it is necessary to keep on top of the torrent of data if businesses want to remain competitive over the medium to long term. 

While few would argue with the need to successfully integrate new volumes and sources of data into their organizations (according to a recent survey by 451 Research, 91% of enterprises say at least some of their strategic decisions are driven by data), the ability to tap data as needed and use it to support informed, agile decision making can be elusive.

In too many organizations, decision makers spend too much time searching through too many data silos to find what they need. To reduce the time required to get at, and use their data effectively, organizations are turning to new technological solutions and methods that help maximize their ability to manage data at the lowest cost possible – and get that data into the hands of as many authorized users as possible. Data lakes and the use of DataOps practices come to mind here.

Ultimately, being data-driven isn’t about a particular technology; it’s about how technology can enhance, amplify and serve the needs of the business by putting the right data in front of the right user at the right time. Assuming the technological components are in place that establish an organization’s ability to seamlessly extract value from their data (see the stairway-to-value process described in the ebook), an enterprise is ready to embark on its data journey.


Encourage collaboration, both internally and externally.

Data is coming in from all corners of your enterprise, but it may often be siloed within departments or be housed on separate systems. To become a digital enterprise, cultural shifts are necessary with new levels of collaboration at the center. Collaboration across the breadth of an organization’s functions, with data-driven decision being the unifying thread, will enable data to be used to help solve business challenges in new ways – and address some challenges that may not have even been recognized previously. But to bring about these changes, data owners and managers need to have an opportunity to collect/manage data from a wide range of sources to see for themselves how collaboration can benefit their function. Many traditional data users are simply unaccustomed to sharing data widely and freely, so demonstrating the value of sharing is a fundamental milestone in the path to becoming data-driven.    
Beyond internal functions, it’s also critical to collaborate and innovate with business and IT/technology partners. In today’s highly networked global economy, no enterprise is an island. Indeed, innovation may come from outside the IT organization or even outside the enterprise.

The emerging practice of DataOps – which is a methodology calling for automation and process improvements to improve data quality and reduce the cycletime of data analytics – provides a framework for building a more collaborative culture. And, just as DevOps leveraged automation for the development and deployment process, DataOps takes advantage of automating the process of getting data from different sources, transforming it, blending it, making it useful and delivering it to the right application or decision maker.

Open data to all levels of the organization.

Once the domain of a narrow group of staff from the IT teams, data was guarded and kept siloed in part by necessity and in part because there were not many users to satisfy. If they are to realize value of the data that surrounds them, enterprises must actively and consistently provide access to data to a much, much broader group of end users, applications and systems. To really derive value, an organization has to create a culture of collaboration and sharing of data in ways that may seem counter-intuitive to some but is at the core of successful DataOps practices. New methods of governance help maintain the levels of data integrity, security and accuracy necessary to develop that culture as broader teams access, blend and analyze data.

But “end users” are not only human – data-driven systems rely on input from many sources and other systems. It’s important to build a “repository of knowledge” that enables the organization to continue improving the effectiveness of data-driven insights. For leading enterprises developing AI-based solutions, this might mean creating a repository to train virtual assistants with a central source of data, which ultimately creates a consistent experience for clients across virtual channels.

Promote transparency.

Transparency helps decision makers understand the lineage and trustworthiness of data. This becomes even more important as data-driven approaches such as AI take hold. AI engines can be seen as opaque, which can raise governance issues. To overcome this, data science and innovation teams need to work closely with risk and compliance teams to build clear audit trails to ensure that humans can understand exactly what actions AI engines take to make decisions.

Always stay focused on the customer.

One of the primary goals of becoming data-driven is to better understand, anticipate and meet the needs of the customer. One way to do this is to start with identifying customer needs and behaviors and working backwards to determine what data – either existing or that needs to be accessed – is available to apply to the effort. By directly connecting the data with the end goals of serving customer needs, organizations can continuously improve by quickly and accurately integrating customer preferences into product/service offerings and using communications tools to intake and acknowledge pain points.

Create a culture around measurement and analytics.

Most organizations exist to satisfy a customer base through delivery of a product or service, but relatively few have been able to make the types of cultural changes that allow them directly respond to the data being returned by their customers.  One example of the power of measurement and analytics is knowing early on whether something is or is not likely to be a success – if customer feedback data shows that a product or solution is not adding value for customers, companies are able to quickly pivot away from those activities and focus on higher value investments. To gather this type of information, companies can use online channels to regularly perform A/B tests that allow rapid, continuous and inexpensive feedback. That data can be used to make successful offerings better and identify opportunities to modify or pull back on offerings that don’t resonate as much with customers.

For an enterprise to gain this kind of sustained value from its data investments, it must create a culture of analytics where decisions are vetted by data, business users are motivated to examine data, and the organization uses data to proactively engage with the market. Similarly, a strong culture of data governance must be fostered in which individuals and groups understand the parameters of self-service data investigation and willingly go through governance reviews when distributing information to a broader audience.

Improved data governance is a big benefit of moving to a DataOps model. For example, one of the most complex and expensive challenges, regulatory compliance and reporting, is complicated by the fact that data sources extend across private and public cloud. Technology solutions that combine powerful metadata tools and automate the tagging process in a single platform radically simplify compliance and reduce costs.

Automate as much as possible.

Volumes of data moving through enterprises are only going to keep increasing, especially as more systems and capabilities are added. The ability to move data automatically through the four key stages of maturity—storing, enriching, activating and monetizing—will enable managers and staff to focus on business opportunities, not on data management and scripting, enabled by advanced analytics, AI and machine learning. By applying these emerging technologies to data, teams can gain more insights into the internal functioning of their business at a lower cost than previously possible.

DataOps: The Necessary Shift to Being Data Driven

As organizations begin their journey becoming truly data driven, the role of data in business success is increasingly being recognized, and best practices such as DataOps— driving collaboration and alignment, and making extensive  use of automation to reduce the cycle time of data analytics, improve data quality and reduce data risk or avoid compliance violations—are emerging to help process and deliver data-driven insights where and when they’re needed. DataOps is enterprise data management for the AI era, with the goal of seamlessly connecting data consumers and data creators to rapidly determine and use the value that is in data of all types. Data operations is not a product, service or solution, it's a methodology that requires technological and cultural changes to improve use of data through better data quality, shorter cycle time and superior data management.

Fear of disruption is driving a commitment to data as digital upstarts free of brick-and-mortar overhead and enabled by modern data infrastructure and digital sales and marketing channels find new ways to deliver value to the customer. A data-driven enterprise is built upon a forward-looking technology architecture an infrastructure that is robust, adaptable and open, and a commitment to cultural change. The commitment to data corresponds with key strategic initiatives around digital transformation, AI, and developing a 360-degree view of customers and suppliers. By competing on data, organizations will have the insights on customers, markets and environments that will provide a competitive edge. The key is to be ready to embrace the data revolution.