Three considerations when building an intelligent data and AI platform

Until recently, data and artificial intelligence (AI) was only discussed, managed and owned by teams of technologists within an organization. Those folks who lived and breathed data and the platforms that housed it.

Not anymore.

Data and AI initiatives have taken center stage as C-suite and even boardroom priorities, as executives recognize the growing value of being a data-driven business. It’s a shift to thinking about how the right use of data and AI can help them uncover significant new opportunities for growth. But the reality is… it’s a lot harder to do than people might think. Most companies still struggle to use their data to transform their business.

The biggest challenge companies face is a proliferation of data across platforms and systems. This makes it very difficult for companies to make data accessible to people in the right form and context—there’s no single place people can go to consume data or get data-driven insights. It also makes it virtually impossible to apply AI at scale across the enterprise to uncover new sources of value. This issue is even more complex for large legacy organizations that have “grown up” in silo’ed operating model where data also grew up fragmented. Over time the people who set it up are no longer here, creating real data lineage issues that are difficult to untangle as companies try to migrate to the cloud.

So what do companies today need to do to make better use of all of this data? One critical part is getting the foundation right. A modern data architecture on cloud and a connected enterprise data and AI platform gives people access to the insights they need—regardless of where the data sits.

This enterprise backbone enables a “one stop shop” or marketplace where people across the enterprise can go to get various data products to help drive the business. Importantly, such platforms not only enable access, but they also provide data in a way that’s contextual to the needs of specific data users and consumers.

Companies serious about using data and AI for growth need to begin taking steps to modernize their data foundation so their people can unlock the immense value that remains trapped in the rich data across their enterprise.

Underpinning this are 3 critical considerations:

Organizations everywhere are awakening to the fact they’re sitting on a treasure trove of data they could be using in so many ways to transform their business. With the right data and AI platform, data users can more easily apply analytical models to a broader set of critical data, thus generating better and more powerful insights about business performance, new growth opportunities, and potentially impactful innovations.

Users who need to make day-to-day decisions about the business can leverage data by “subscribing” to relevant insights and having those insights delivered to them on a regular basis. This could be, for instance, a weekly refresh of specific KPIs they need to monitor, such as foot traffic in a particular retailer’s area or weather patterns that could affect the supply chain. Or it could be delivery of up-to-the minute information on how a specific manufacturing line is operating and suggested actions a plant manager should take to improve the line’s performance. Or it could be a steady stream of real-time insights on how pricing is affecting demand for products in a certain city and what the optimal prices should be at any given moment to generate the greatest sales. It’s creating a “data supply chain,” bringing together data from multiple sources (including databases, devices, and sensors), curating it, and making it available for use.

This is key to acquiring all the relevant data users need, cleansing and standardizing the data, and preparing the data to be accessed and manipulated by users.

A common misconception is that you have to wait on generating business value or getting business teams integrated into the development of the data foundation from the onset. It’s imperative that key stakeholders in both the business and data/IT teams are involved in defining what value looks like, aligning on priorities, are aware of critical milestones and have a sense of ownership throughout the transformation.

Data landscapes are only as good as the management and governance of the data itself, which can vary significantly from system to system. At the same time, they must take a proactive approach to ensure data and AI are used ethically and responsibly to avoid negative consequences and violating customers’ trust. It takes enormous effort and expertise to manage and govern all these different landscapes consistently to 1) ensure data quality and veracity (so users trust it) and 2) provide a simple, intuitive access experience for those who need to use the data (or insights derived from it), regardless of where it sits. Because most companies don’t have the resources or skills to do this, data quality and the access experience can vary substantially by system and data set—ultimately hindering efforts to unlock data’s value.

Data sharing is an important KPI and a business necessity. It accelerates digital business transformation and more than ever, leaders are responsible for ensuring data quality, creating security measures that anticipate threats and building ethically responsible frameworks for managing data and AI. This establishes a virtuous cycle of data creation and consumption, because quality is always improving. Data practitioners don’t rely on unverified or biased data to make decisions and instead take steps such as using AI itself to detect biased algorithms.

With data security, controls and governance processes in place, the organization can delineate who can access which data (or data-related products) and how they can use it. There is a delicate balance between protecting critical data from misuse or being exposed to those who aren’t authorized to see or use it, with making it useful to those within the organization to improve the business. Management and governance rigor is needed to ensure that all data is consistently of the highest quality across all systems and platforms, and that data is fully traceable from when it’s acquired to when and how it’s used. This is critical to building user trust in the data and the insights generated from it. A comprehensive data rights management solution enables data creators or producers to retain control of their data products and Consumers to have a seamless experience in accessing data products.

Creating a positive user experience is critical for adoption of your data and AI strategy. Each part of the organization may be looking at the same data but through a different lens.

Producers of data products go to multiple data management systems to define and manage their products. With each function offering a potentially different perspective, you get different offshoots of data products which lack consistency and can cause confusion when reporting.

Having the right tools that enable data to be consumed and used effectively and in a way that’s meaningful and relevant to individual users is key. This could be a data marketplace, a sandbox workbench, enterprise data search functionality, business intelligence reporting, and data visualization platforms. For example, a data marketplace or data platform enables reusability of data products, increasing efficiency and value of data projects. This is key to creating a superior user experience that encourages people to access an apply data in ways that support their role. It also evangelizes the use of data and insights for decision-making across the enterprise.

Improving the user experience is also about efficiency – having simpler and more intuitive processes to build, manage, consume and use data reduces the time to perform data-related activities, reducing the time to insights.

Building the right capabilities are a necessary precursor to scaling data and AI usage and enterprise-wide adoption because it 1) builds confidence/trust in the data itself and 2) designs data systems around business users/their needs versus creating new complex hoops for business users to jump through if they want to use data.

It isn’t just about making tools and resources available though. Finding and retaining people skilled for roles in data and AI is currently still a big limitation for many organizations. In a recent Accenture study of Chief Data Officers, responses indicated that lack of talent to operationalize, and slow ability to change/adoption as the two biggest barriers to transformation. For optimal user adoption, the foundation needs to enable end users of all technical skillsets and abilities to use data effectively and efficiently.

This means creating programs that improve data literacy, training on new and emerging technologies and creating programs which prepare employees for the roles of the future.

Whether you’re in the early stages or well on your way to building your data and AI platform, it’s never a bad time to pause and reflect on how you can ensure what your building aligns to your business goals and is built to encourage user adoption.

The fact is, companies that lead with data for decision-making are differentiating themselves. They use insights to understand customers more deeply and transform the business to serve them better. And they can uncover new opportunities and exploit them faster than competitors. It all adds up to faster, more robust growth. Accenture research shows that companies that have embraced data and AI boost their revenues five times more quickly than those that haven’t.

With a modern enterprise data and AI platform and foundation paired with the right data culture, organizations can unlock more value, better understand and connect with customers, operate far more efficiently, and innovate in ways never before possible.