Challenges of using artificial intelligence and adoption
Data is constantly in motion—moving quickly from person to person and from person to machines and back. As many AI-fueled organizations can attest, the magic happens when data is transformed into value, even profit, enhancing workforce and customer experiences alike. But many organizations struggle to capture and manage it to their business advantage.
In Deloitte’s third annual State of AI in the Enterprise Survey, companies that have adopted AI at some level reported challenges in critical aspects of data management. These challenges included integrating data from diverse sources, preparing and cleaning data, providing self-service access to data, ensuring data governance, and lacking the right talent and expertise to manage the data value chain. In fact, at least 40% of AI adopters reported a low or medium level of sophistication across a range of data practices. And nearly one-third of executives identified data-related challenges among the top three concerns hampering their company’s AI initiatives.
Building an effective data management value chain can lead to powerful and game-changing benefits. Forward-looking data-driven companies are bringing in a product mindset, managing the data like a product across its entire life cycle.
AI is now more pervasive
AI applications are soon to be everywhere, and organizations are already adopting the technology at varying levels. Enterprises are ambitious in using AI to disrupt business models for competitive advantage and value creation for stakeholders. These organizations recognize the business imperative of maturing to become AI-fueled. As more companies experiment with AI, advance their data-related capabilities, acquire new technologies and talent, and integrate AI into their business processes, they are facing inherent challenges in data management. Data and AI share a powerful connection.
Setting the pace in AI adoption maturity
of organizations are “seasoned” at AI adoption.
of organizations are “skilled” at AI adoption.
of organizations are “starters” at AI adoption.
Powerful potential, but significant challenges
For AI to succeed, organizations should address data challenges and fix bad data, applying principles to better manage, clean, and enrich it so broader AI ambitions can be met. But most haven’t reached a level of maturity in data management capabilities, and about a third of AI programs fail as a result.
While vast amounts of data are available to organizations, it is rarely interconnected or integrated to realize its benefits. This hurdle can make it more difficult for organizations to leverage not just their own internal data but data from external sources. In addition, important insights can be missed due to lack of complete or standardized data, and this can produce inaccurate analysis and reports.
From a resource perspective, too much employee time—and associated expenses—are going toward managing and preparing data for one-off analysis. Contributing to that is a lack of data talent trained on new and innovative methods of managing large data sets.
The larger picture reveals that a data culture is missing from many organizations’ mindset, and responsibility for good data has not been adopted at the enterprise level. These challenges can lead to the risk of unintended consequences such as AI failures and unanticipated results. Organizations should know how to select the right data to reduce or eliminate biases in their models.
Changing the game
AI adopters are reimagining, disrupting, and transforming the data value chain. Such adopters ranked modernizing data infrastructure for AI as the number-one focus for their AI initiatives. They realize their organizations can’t effectively implement AI without a modern, robust infrastructure. Once the data management infrastructure is modernized at an enterprise-wide level, business units aren’t having to constantly reinvent the wheel. Data becomes democratized.
How well an organization can integrate and ingest data; standardize, cleanse, and curate the data; and ultimately consume its data can determine its future success. Organizations that can fuel AI with data and computing power can turn their investment into both value and profits. Companies that undergo a robust data modernization effort, often migrating it to the cloud, can serve up insights that delight customers and transform workforces
Without a robust, repeatable, and scalable data value chain, AI can’t scale. It’s not the single AI use case but hundreds of use cases that create game-changing value for organizations. That’s when the possibilities can really open up. Organizations can ingest data, transform it, drive insights, and execute business processes at a faster pace with more accuracy than they ever have before. For AI-fueled organizations, data becomes a resource, sparking innovation and competitive advantage.
Building a transformative data organization
To help be truly transformative, organizations should have a bold, enterprisewide strategy that is established and championed by the highest leadership. Three strategic levels to an innovative program include:
- What’s at the core of a data-fluent program?
- How is data delivered in a connected world?
- How is the program impacting the business? Each level carries its own components or capabilities
Organizations spend a lot of time on data readiness and platform-related capabilities. However, without a strong data-first culture at the core, it can be impossible to drive innovation and value. Data and AI are connected. AI delivers the insights that help create value, but it can’t do it if the data architecture isn’t aligned to the business from the onset. Also, irrespective of the cutting-edge data capabilities put to use, business stakeholders typically measure the performance of the data organization based on how well they engage with the business and what value they ultimately realize. Essentially, it doesn’t matter how good the data technology is, if the business doesn’t recognize how to use the asset. Data turns to value when the business can draw upon AI-fueled insights and take action in the moment.
Level 1: Strengthen the core
An enterprisewide data-first culture should be in place for an effective use of data to occur. Part of this involves building advanced data capabilities, but it’s also raising a base level of data literacy across different levels of the organization. The magic of innovation can happen when people start reimaging their own day-to-day work and realizing the benefits that quality data can bring.
For that to happen, data readiness is key. Quality, free-flowing data takes effort. The companies that use advanced technologies and even AI to compress time frames and address volume challenges can make the job a lot easier. To help hit that North Star, companies should spend 20% to 30% of their time managing data, and they can only hit that target by automating processes.
Types of data automation tools
Level 2: Enhance delivery
To help connect data, a modern data platform on the cloud can be used to ingest and curate real-time information. It should scale, flex, and support a range of systems, applications, and users. A successful modern data platform minimizes effort, improves accuracy, and speeds up time to delivery. Security and privacy initiatives are a key part of the effort as is a clear data governance process that ensures data is trusted and risks such as biases are minimized. The addition of a digital workforce provides an opportunity to retool positions, upskill people, hire the right AI talent in terms of roles and skills, and rely on outsourcing when needed. New capabilities require new processes not just across the organization, but also within IT and data science teams. It also necessitates a robust ML Ops strategy to realize implementation goals, ongoing quality, and ethical delivery.
From there, it’s the cloud data analytics that power new innovative thinking as employees have more relevant insights faster than they ever did before.
Level 3: Impact the business
The engagement level should inspire business functions to continuously access data so it can be useful in developing new solutions to problems and accelerating value. Insights from free-flowing data and AI result in value for the company as it achieves business outcomes at new levels. At the end of the day, data modernization can help spark the magic of turning data into profit:
Revenue growth
Data can help acquire new customers and retain existing customers by providing insights to strengthen pricing strategies, improve cross-selling services, and better manage supply and demand.
Operational efficiency
Automating data tasks can save time for busy data engineers and business analysts. Because all departments need access to data, creating a repeatable framework to connect data sources provides a cost advantage to those using modernization techniques.
New opportunities
Data modernization platforms open up new opportunities for companies by being able to share and even monetize their data with broader ecosystem partners.
The path forward
An AI-first data value chain can allow the organization to better ingest data, transform it, drive insights, and execute business processes at a faster pace and with more accuracy. Some companies may be eligible for certain R&D tax credits that can help offset some cost as well.
Going forward, AI-fueled organizations can move faster because adding new, higher-quality data sources can yield game-changing benefit across the enterprise as essential insights reach people in time to really make a difference for their customers and key stakeholders.
Get in touch
