Researchers identify scale, ecosystem leverage and premium positioning as sustainable competitive paths.
A newly published study in California Management Review challenges a persistent assumption in today’s AI landscape: that the strongest model inevitably becomes the dominant platform. Instead, the authors of “Three Winning AI Strategies” show that the global AI assistant market is being shaped by three distinct strategic positions—each succeeding on its own terms.
The study—by Daniel McCarthy of the University of Maryland’s Robert H. Smith School of Business, Maxime C. Cohen and Eddy Hage-Youssef of McGill University, and D. Daniel Sokol of the University of Southern California—draws on worldwide mobile app usage data through December 2025.
The researchers show that ChatGPT, Google Gemini and Claude occupy fundamentally different competitive positions: a scale leader, an ecosystem player and a premium niche, respectively. “So many people are so focused on which AI company has the best model, but there is much more to the competitive market than that,” says McCarthy.
Three Models, Three Ways to Win
Sensor Tower estimates, the authors note, reveal stark differences in how the leading assistants participate in the market:
- ChatGPT — Scale Leader. With 78.4% of worldwide mobile daily active users, ChatGPT dominates in reach. Its challenge is converting that massive user base into durable economics.
- Gemini — Ecosystem Player. Holding a 12.5% share of mobile daily active users (DAU), Gemini generates little direct app revenue. Instead, its value is plausibly captured across Google’s broader ecosystem of complementary products.
- Claude — Premium Niche. With 0.5% of mobile DAU share, Claude operates at small scale but with striking economics: estimated revenue per active user more than 40x Gemini’s and roughly 3x ChatGPT’s. Its strategic test is maintaining differentiation as it grows.
“These differences highlight a central finding,” McCarthy says. “ChatGPT is selling to everyone. Google is giving Gemini away to protect its search business. Anthropic is charging a small group of professionals real money. All three are working for now, but they’re working for completely different reasons.”
From a consumer standpoint, the reason these assistants feel so different is straightforward: they’re intentionally built for different kinds of users. “Gemini is built to keep you inside Google. ChatGPT is built to be the thing everyone tries first. Claude is built for people who don’t mind paying for it as much,” McCarthy says. “That’s healthy. A market where every product felt the same would be a worse market.”
Major Launches Expand the Category — Not Cannibalize Rivals
If AI competition were a simple “horse race,” major model launches would trigger visible declines among rivals. The study finds the opposite. Across 15 major releases between May 2023 and December 2025:
- Competitor effects are mixed and centered near zero (mean −0.1%).
- Focal‑firm effects are meaningfully positive (mean +7.0%).
- Total market usage expands substantially over the period.
“In other words,” McCarthy says, “launches help the launching firm without systematically hurting competitors. If this were a horse race, every big launch would knock the other horses back. We looked at 15 launches, and you don’t really see that.”
Different Strategies, Different Economic Tests
As the market matures, each strategic position faces a distinct pressure point:
- Scale: Can a massive user base be monetized sustainably?
- Ecosystem: Can value be captured across complementary products when the standalone service is inexpensive or free?
- Premium: Can distinctiveness—and willingness to pay—survive as the audience broadens?
Market share alone does not determine strategic success, McCarthy says. “What matters is whether a firm chooses a position aligned with its structural advantages.”
The authors conclude that the strategic error is not choosing the “wrong” universal model—it’s pursuing a position that doesn’t fit the organization’s strengths. Executives must ask:
1. Scale: Do we have a credible path to serving and monetizing a very large user base?
2. Ecosystem: Can our distribution or complementary offerings capture value the focal product does not?
3. Premium differentiation: Can we solve an important problem well enough to command higher willingness to pay?
Ultimately, the firms that endure will be those whose strategies match their structural advantages—not those chasing a single dominant model.
