Earnest Analytics: ~60% of paid Gemini users kept their subscriptions six months after signing up, better than Perplexity but lagging behind ChatGPT and Claude
Tired of trailing OpenAI, CEO Sundar Pichai wants his Gemini chatbot to be used by 500 million people
Context & Ripple Effects
Gemini’s paid-subscriber durability offers an early quality-of-demand measure alongside Google’s stated ambition to expand the chatbot’s audience. It also establishes a baseline for a product that later ranked second to ChatGPT on mobile in a consumer-AI-app survey, at roughly half of ChatGPT’s monthly active users Gemini’s No. 2 mobile position.
Subsequent coverage describes a more competitive consumer chatbot market: Gemini’s monthly active users grew while ChatGPT remained far larger Gemini’s user growth alongside ChatGPT’s scale, and later market-share estimates showed the gap narrowing. Retention helps distinguish recurring paid use from headline-grabbing adoption.
First-order effects
- Gemini has evidence that a majority of paying subscribers still see enough value to remain after six months, though its retention trails ChatGPT and Claude.
- The comparison puts immediate pressure on Gemini to improve paid-user stickiness relative to the two higher-retaining assistants, while giving it a stronger retention position than Perplexity.
Second-order effects
- Competitors will be judged less on acquisition and more on whether their paid tiers sustain repeat use; Gemini’s later reported rise in chatbot share Gemini’s growing chatbot share makes that benchmark more consequential.
- Google can evaluate its push toward broader Gemini adoption against a clearer monetization signal, while users deciding among subscriptions gain a comparative indicator beyond model capability claims.
Third-order effects
- Consumer AI competition is likely to shift toward the combination of distribution, engagement, and paid retention rather than raw user counts alone.
- If retention gaps persist, the leading assistants may gain a compounding advantage in recurring revenue and product iteration; if they close, distribution channels could matter more in determining which subscriptions scale.
The trend: Consumer AI is moving from an initial adoption race toward a contest over durable paid usage and the distribution systems that support it.