Microsoft says AI’s real test is execution, not experimentation

Microsoft is putting a line in the sand for enterprise AI: the era of polished demos is giving way to proof. In a May 21 blog post, the company said organizations are moving from AI experimentation to enterprise-scale transformation, and that leaders are now prioritizing measurable outcomes, faster time to value and repeatability across the business.

For monday.com, that message lands close to home. The company has spent the past year recasting itself from a work-management platform into a work-execution platform, first with monday magic, monday vibe and monday sidekick, then with monday agents and monday campaigns. Chief Product & Technology Officer Daniel Lereya framed that shift bluntly in September, saying software was entering an era where it does not just manage work, it does the work.

That is more than product language. It raises the bar for the people building and selling the platform. Engineers now have to deliver reliability, permissions and deeper integrations, because AI features will be judged by whether they hold up inside real company systems. Product managers have to measure success by workflow impact, not novelty. Sales teams have to prove that AI can scale across departments, not just save one team a few minutes.

monday.com has been backing that pitch with numbers. In fiscal 2025, the company said revenue reached $1.232 billion, up 27% year over year, and customers with more than $50,000 in annual recurring revenue made up 41% of total ARR. The company also said monday vibe was the fastest product in its history to top $1 million in ARR, a signal that AI is already becoming part of the commercial engine, not just the roadmap.

The upmarket push is visible in other operating metrics too. In the second quarter of 2025, monday.com reported revenue of $299.0 million, up 27% year over year, and said net dollar retention among customers with more than $100,000 in ARR was 117%. By the first quarter of 2026, the company said it had more than 250,000 customers worldwide. That mix suggests the platform is trying to win larger, stickier accounts where AI has to earn its keep in finance, operations and customer-facing workflows.

Security is part of the pitch as well. monday.com’s AI Trust Center says customer data is encrypted in transit and at rest and is not used to train machine-learning models, a safeguard that matters as buyers grow more cautious about what gets fed into AI systems. For a company that has spent years selling flexibility, the new test is whether that flexibility can translate into measurable throughput. In a market crowded with copilots and agent demos, the winners will be the platforms that can show real work getting done.