Successfully implementing AI tools across a company is one thing. Getting consumers on board is entirely another—and according to a recent report, more than 97% of companies are missing the mark.
Global communications firm Gregory just released a new study on company’s AI strategy rollouts, revealing that companies with better executed announcements see massive differences in returns on the stock market compared to those that phone it in.
How does one quantify the quality of an AI strategy rollout? Gregory measured companies across five key dimensions to give each an AI Communications Quality Score (ACQS) following their announcements. From 2022 to 2025, Gregory assessed 449 companies.
Those key factors include CEO ownership, or how personally connected and involved a CEO is in their company’s AI rollout; named use cases, with companies highlighting specific ways they’ll apply AI rather than vague promises to explore; 90-day follow-through, assessing how well a company put its strategy into practice in the three months following its announcement; board and governance record, or how a company’s board has previously demonstrated a commitment to AI oversight across its institution; and tier-1 media coverage, with the announcement getting picked up by credible outlets.
That adds up to a score between 0 and 20. Gregory further broke down its 100 highest-scoring companies into three tiers: Tier 1, with scores 18–20; Tier 2, with scores 15–17; and Tier 3; with scores 12–16. Of the 449 companies scored, only 13 scored in the highest tier, less than 3% of the total sample.
In the 90 days after their announcements, companies in Tier 1 saw an average alpha improvement of 10.8% compared to sector benchmarks. Meanwhile, companies in Tier 3 lost 2.2% alpha on average over the same period. That’s a 13-point gap between the strongest communicators and the weakest.
Companies in Tier 2 also saw an alpha increase on average, but it paled in comparison to Tier 1’s: Tier 2 companies gained just 1.2%, nine times less improvement than Tier 1.
Not just titans of the tech world
Though some of the highest performers across the study were titans of the tech world, including Alphabet, Meta, Amazon, and Microsoft, the pattern holds true even ignoring their scores. Excluding those companies along with Apple and Nvidia, Tier 1 companies still outperformed Tier 3 companies by 11.3 percentage points.
“The biggest difference we saw across AI announcements was the level of commitment and follow-through,” Greg Matusky, CEO at Gregory, said of the firm’s findings. “Every top scorer had a CEO who owned the story personally, named specific use cases with real business stakes, and shipped something within 90 days of the announcement. The lower tiers left AI to an earnings call conversation or a press release with vague plans, and the market treated those announcements as noise.”
What should companies make of these results? Gregory recommends companies follow a pre-announcement checklist, making sure they can answer “yes” to questions like, “Is the CEO delivering this personally?” and, “Is the 90-day follow-through already scheduled?” before releasing any news about new AI initiatives.
“The data suggests waiting costs less than announcing badly,” reads the study. “The weakest announcements in this study correlated with negative excess returns, which means a poorly told AI story performed worse than saying nothing at all.”