Chatbot wars: OpenAI growth slows but still leads rivals in total usage, says Sensor Tower

ChatGPT's monthly average user growth in May lagged competitors, even though its overall user base remains well above rivals, according to Sensor Tower data. OpenAI's ChatGPT MAUs are still up 58% year over year, indicating continued growth despite a slower recent pace. The report is informative rather than event-driven and is unlikely to move the stock meaningfully on its own.

Analysis

Slowing relative user growth at the category leader matters less for near-term revenue than for the narrative premium embedded across AI infrastructure and app-layer beneficiaries. The important second-order effect is that the market may begin to distinguish between distribution wins and monetization wins: if the incumbent is already saturated on engagement, incremental upside shifts from consumer app names to pick-and-shovel exposure in compute, data-center networking, and model deployment. That tends to favor the infrastructure complex over the application layer, where usage growth alone is not enough to justify multiple expansion.

The competitive read-through is more nuanced than a simple share-loss headline. A high absolute user base with decelerating growth often signals the start of a transition from land-grab to retention, which can be healthy if it coincides with better paid conversion; if not, it increases the risk that rivals can outgrow the leader by targeting specific workflows, geographies, or enterprise use cases. Over the next 1-3 quarters, the market will likely price every product launch, default placement change, and distribution partnership more aggressively because the bar for sustaining premium sentiment is now higher.

The main tail risk is that investors confuse relative growth deceleration with terminal weakness. If the incumbent is still compounding from a much larger base, the correct competitive benchmark is monetizable attention, not raw MAU growth; that creates room for the leader to preserve pricing power even as headline growth normalizes. The contrarian angle is that slower growth can actually be bullish for margins if it reduces acquisition intensity and improves infrastructure utilization, especially if paid users grow faster than free users.

For traders, the best setup is to lean into the infrastructure trade while fading indiscriminate app-layer enthusiasm. The key catalyst over the next 30-90 days is evidence that usage is translating into paid conversion and enterprise attach rates; absent that, multiple compression is more likely than outright revenue disappointment. If growth re-accelerates after feature releases or a new distribution deal, the move should reverse quickly, so positions need to be paired and time-boxed.

Key Decisions for Investors

• Long AI infrastructure basket vs. AI app-layer basket over 1-3 months: favor NVDA, ANET, SMCI over high-multiple consumer AI names; use a relative-value pair to express the view that monetization migrates upstream.
• Short momentum in unprofitable AI software names for 4-8 weeks if they trade primarily on usage narratives; set a tight stop in case of a surprise distribution partnership or paid-conversion update.
• Pair trade: long MSFT / short a basket of AI consumer-application exposure over 1 quarter, on the thesis that enterprise distribution and embedded workflow integration are more defensible than standalone chat growth.
• Buy upside in compute beneficiaries via call spreads in NVDA or ANET into the next earnings cycle; risk/reward improves if the market rotates from user-growth headlines to capex and utilization data.
• If you must express a view on the category leader, prefer a call spread rather than outright long exposure: the upside is tied to monetization conversion, while the downside is that growth normalization compresses the premium quickly.