AI Delivers Productivity Gains Yet Fails to Unlock Revenue Growth, Harvard Survey Finds

A new survey of 385 business decision-makers reveals that while artificial intelligence has moved beyond the pilot phase in most organizations, the technology is delivering incremental productivity gains rather than the top-line growth executives anticipated. The research, conducted in March 2026 by Harvard Business Review Analytic Services and sponsored by process automation firm Appian, found that 59 percent of organizations now operate AI in production environments, yet the majority remain focused on efficiency improvements that fail to translate into measurable revenue expansion.

Of the performance metrics organizations track, 64 percent reported productivity improvements and 58 percent cited operational efficiency gains. By contrast, only 30 percent saw impact on new revenue streams and just 35 percent measured improved return on investment. The disparity suggests that enterprises are deploying AI to automate existing workflows rather than reimagine business models or create new market opportunities.

The findings align with separate warnings from technology services providers about a widening skills gap that threatens to undermine AI ambitions. Hexaware, a global IT solutions firm, cautioned that UK businesses risk falling behind in what the government estimates will be a £400 billion AI economy by 2030. The company pointed to a shortage of readily available talent and recommended that organizations turn to technology partners and Global Capability Centres in emerging hubs to supplement internal teams with specialized expertise in data science and AI engineering.

Meanwhile, evidence of AI integration into daily work routines continues to accumulate. Teachers are using the technology to create lesson plans and grade papers, while marketing professionals at staffing firm HireQuest have built dashboards that analyze website traffic and social media trends to inform franchisees. Ashley Smith, head of marketing at HireQuest, described uploading screenshots from a manufacturing trade show to an AI platform, which then generated a list of company names and predicted staffing needs based on press releases and stock reports.

Yet the enthusiasm for deployment has collided with mounting concerns about trust and accuracy. Dr. Chih-Han Yu, Chief Executive Officer and Co-Founder of Appier, said that "the central question for enterprises has shifted from whether AI can be used to whether it can be trusted." He noted that many AI systems deliver confident responses even when uncertain, creating operational risk that extends beyond poor user experience to customer service errors and content hallucinations. Appier announced research capabilities aimed at positioning agentic AI as a trustworthy decision partner, emphasizing that "the true source of enterprise advantage will be whether AI can be trusted to make decisions."

Duane Barnes, president of RapidScale, a Cox Business company, emphasized that companies succeed most when they first define the problem AI should address. "AI doesn't create value just because it's deployed—it has to be connected to clean, trusted data and integrated into the workflow where decisions actually happen," Barnes said. RapidScale recently helped Barrett Financial Group scale operations from a few hundred loans per month to more than 3,000 by automating initial processing, while Cox Business reported a 35 percent increase in average deal size after unifying targeting, marketing, and sales workflows in a pilot program.

(The Harvard Business Review Analytic Services survey was sponsored by Appian, a Nasdaq-listed process automation vendor. Separate announcements from Hexaware, Appier, and RapidScale were issued independently and reflect each company's commercial positioning in the enterprise AI market.)

The productivity-versus-growth divide mirrors broader industry patterns documented by research institutions. MIT found that only about 5 percent of enterprise generative AI pilots deliver measurable financial impact, while Gartner forecasts that 40 percent of agentic AI projects will be scrapped by 2027 due to escalating costs and unclear business value. The gap between deployment enthusiasm and revenue realization suggests that many organizations have automated processes without rethinking the strategic architecture required to capture market share or launch new offerings. As AI systems become more capable of executing tasks, the challenge has shifted from technical feasibility to organizational design and the ability to align automation with growth objectives rather than cost reduction alone.