During the ChatGPT strike, Google Gemini took advantage and saw a 60% surge in search volume

Author: Fang Jiayao

Source: Hard AI

In the early hours of June 5th, OpenAI's ChatGPT suddenly "went on strike" due to a system failure. During the interruption of ChatGPT service, a large number of users worldwide sought alternatives and flocked to other AI platforms, causing a surge in traffic that triggered a domino effect, leading to the consecutive crashes of large models such as Gemini under Google, Claude and Perplexity under AnthropicAI.

According to analysis by QR Code Generator, within four hours after the initial failure of ChatGPT, the search volume for Google's AI chatbot Gemini surged by 60% to 327,058 times.

Global search volume for Gemini soared to 327,058 times

Furthermore, the analysis team pointed out that the growth in searches for "Gemini" was highly correlated with the search trend for the keyword "ChatGPT failure," indicating that users viewed Gemini as a direct alternative to ChatGPT.

The search trends for "ChatGPT failure" and "Gemini" on June 4th showed similar patterns

Normally, Gemini's global daily search volume is 204,991 times. However, during the interruption of ChatGPT service, the demand for Gemini searches sharply increased. In addition, the search volume for "Microsoft Copilot" and "ChatGPT alternatives" also rose, but compared to Gemini, the search volume for these keywords remained relatively low. The search volume for "Microsoft Copilot" increased by 50% to 16,677 times, while the search volume for "ChatGPT alternatives" surged from an average of 898 times per day to 11,000 times within the first few hours.

The search volume for "Microsoft Copilot" and "ChatGPT alternatives" remains relatively low compared to Gemini

Similarly, reports from DownDetector also showed that ChatGPT encountered its first failure before 3 p.m. Beijing time on June 4th, during which period, the search volume for Google's AI chatbot Gemini saw a significant increaseNote that ChatGPT experienced two malfunctions. The first malfunction occurred before 3:00 pm on June 4th, Beijing time. The second malfunction happened at 10:33 pm on the same day. It wasn't until 1:17 am on June 5th that OpenAI announced that all systems had returned to normal operation.

The image shows around 8:00 am London time, corresponding to 3:00 pm Beijing time