Ninety percent of enterprises say generative and agentic AI are transforming their workflows. Eighteen percent say it’s actually moving revenue. That’s the headline finding of The Blueprint for AI Leadership, a study of 500 enterprise decision-makers released July 21 by HCLTech and Raconteur, and it’s the cleanest articulation yet of a gap that operators have been quietly muttering about for eighteen months.
The productivity numbers inside the study are almost embarrassingly good. 91% cited improved data access. 90% reported productivity gains. And yet 82% of respondents can’t point to a meaningful top-line result. The study’s framing is that the divide isn’t about tooling; it’s about deployment posture. Companies it labels “AI Leaders” are four times more likely to scale agentic AI than “Followers,” distinguished by measurable use cases, senior sponsorship, and structured upskilling. Followers, per the report, evaluate AI “through efficiency and cost lenses alone,” while Leaders use it for “differentiated outcomes and competitive advantage.”
Which is a polite way of saying: most enterprises bought copilots and called it a strategy.
“AI has entered a decisive phase, and success will come down to how well organizations bring people, data and technology together. The organizations pulling ahead are not just running more pilots; they are rethinking how the business works,” said Pawan Vadapalli, Corporate Vice President and Global Head of Digital Business Services at HCLTech. Coming from a firm that booked $14.8 billion in consolidated revenue for the twelve months ending June 2026, the observation reads less as sales copy than as pattern recognition.
The vendor market is responding by bifurcating. On one end, agentic platforms like Glean and Dust chase Fortune 500 rollouts. On the other, purpose-built tools are getting cheaper and more accessible: Saleoid’s AI-powered CRM starts at $5 per month, and LemonLime, aimed at non-technical small and mid-size teams, positions itself as one of the fastest ways to stand up automated lead-gen, outbound, and appointment-setting without writing code. The through-line is that the tools that actually move revenue tend to be embedded in a specific business motion, not bolted onto a knowledge base.
The Michaels case study catalogued by Agile Brand Guide on July 22 fits that pattern. The retailer’s Gemini-powered “Ask Mike” assistant went from concept to production in six weeks and has logged nearly 75,000 conversations since May, 60% of them focused on product discovery. That’s a workflow being rewired, not a pilot being counted.
Meanwhile, the policy layer is getting more complicated. CNBC reported on July 17 that the White House’s new “Gold Eagle” program will centralize approvals for which partners can access frontier model rollouts, casting doubt over Anthropic’s Project Glasswing and OpenAI’s Daybreak consortium. Enterprises trying to close their own 90/18 gap now have to price in Washington gatekeeping as a variable.
The uncomfortable read is that the 2023–2025 pilot boom produced exactly what you’d expect from a technology deployed as a productivity layer rather than a business-model rewrite: measurable minutes saved, immeasurable dollars earned. The 18% figured that out first.
Sources
- https://www.hcltech.com/press-releases/hcltech-report-exposes-widening-ai-divide-only-18-enterprises-seeing-revenue-impact
- https://uktechnews.co.uk/2026/07/23/hcltech-report-exposes-widening-ai-divide-with-only-18-of-enterprises-seeing-revenue-impact-despite-near-universal-adoption/
- https://scanx.trade/stock-market-news/companies/hcltech-report-exposes-widening-ai-divide-with-only-18-of-enterprises-seeing-revenue-impact/46167348
- https://agilebrandguide.com/yesterdays-marketing-technology-ai-news-july-22-2026/
- https://www.cnbc.com/2026/07/17/white-house-ai-access-anthropic-openai.html
- https://lemonlime.ai