By Sam Pierson, CTO at Qlik
The adoption of AI has brought a boost in productivity across organizations. This wave of AI investment has not automatically produced distinctive business applications. Many organizations have gravitated toward the same narrow set of use cases: summarizing meetings, accelerating coding, automating tasks and streamlining workflows. The results of these activities have been compelling. In McKinsey’s 2026 global AI survey, eight in 10 respondents say AI has improved their own productivity, and 60% expect their organizations to increase AI investment over the next year.
However, broad adoption hasn’t consistently translated into enterprise-wide financial impact. Just 6% of respondents qualify as AI high performers: they attribute at least 5% of earnings before interest and taxes to AI and report significant value from its use. And despite AI’s wide availability, few businesses have embedded the technology deeply enough to reshape how they work. AI is being added to established workflows to save time, automate repetitive tasks and reduce manual work. These projects are relatively straightforward to implement and show clear productivity returns. However, the same simplicity that makes them appealing also makes them easy for competitors to replicate.
Efficiency matters, but it is only part of AI’s value. The bigger opportunity is using AI to rethink how work is done in the first place and build new operating models around it. That includes making valuable work possible where it was previously too costly or time-consuming.
Where the real value of AI lies
Companies getting the strongest results from AI are far more likely than their peers to have fundamentally reworked their workflows. Nearly three-quarters have done so, up from 55% a year ago, compared with just one-quarter of other organizations. For these organizations, AI isn’t just one more tool in their tech stack. It’s changing how they make decisions, deliver services and generate value. That might mean moving from answering support tickets faster to resolving problems before customers report them, or trading scheduled reviews for decisions made in real time.
Organizations should start by identifying decisions constrained by delay, incomplete information or the cost of analysis. This includes where employees spend their time sorting through information and where services respond to events, instead of anticipating them.
For a manufacturer, this could look like using AI to combine equipment readings with maintenance history and production priorities, flag equipment that’s likely to fail and recommend maintenance ahead of a breakdown. Alternatively, an online retailer might use real-time data and AI agents to tailor product suggestions and pricing, creating a better experience for shoppers. And a financial services company could delegate routine analysis and monitoring to AI so its people can spend more time on clients, risk and strategy.
Rethinking the business this way is more demanding than layering AI onto existing processes. It requires investment, organizational change and willingness to challenge long-standing ways of working. The potential payoff, though, is an advantage that competitors can’t easily copy.
What is holding companies back
Access to AI models is only part of the equation. Rebuilding a business around AI depends on having solid foundations in place: reliable data, shared business definitions, sound analysis and governance that connects insight to accountable action. This means improving visibility across data sources, establishing clear ownership of critical business information and putting frameworks in place to monitor how AI systems are used and how decisions are made. Those foundations should work across models and platforms, so changing tools does not mean rebuilding trust.
For many organizations, model access is not the main obstacle; usable business context is. All too often data is siloed across systems with little visibility into customer journeys and business processes. The result is AI running on fragmented information. And we know that faster processing does not necessarily mean better decisions. General-purpose AI tools are within everyone’s reach. What sets a company apart is applying AI to its own enterprise data, customer knowledge and operational insight in ways competitors cannot easily replicate.
That makes trust and governance critical to any AI transformation. Companies should begin by pinpointing the data behind their most important decisions, fixing problems with quality and access, and making someone clearly responsible for how that data is managed. That responsibility should extend to what AI may do, when people must approve an action and who owns the outcome. Building this trust depends on business leaders, data teams, technologists and governance functions working together. When ownership and accountability are clear, companies are in a much stronger position to build on efficiency gains and use AI to genuinely transform their business.
The next phase of AI adoption
AI has already proven it can make organizations faster. The next phase will test whether it can make them different.
Getting there requires strong data foundations, disciplined governance and close collaboration across business, technology and data teams, along with a clear view of what the business should look like three years from now. Start with one consequential workflow, assign an owner and measure business impact against the full operating cost, including integration, oversight and rework. Scale what demonstrably works.
