Google is losing search share to LLMs like ChatGPT and Perplexity, but the shift is gradual rather than catastrophic. Google held approximately 89-91% of global search queries as of 2026, down from a peak of roughly 93% in 2021, with ChatGPT and Perplexity absorbing a growing slice of informational queries. Web SEM tracks this shift closely to help clients maintain visibility across both traditional search and AI-powered answer engines.
The decline is real but measured. Google still processes an estimated 8.5 billion searches per day, while ChatGPT handles over 1 billion queries per day and Perplexity AI reported approximately 100 million queries per week in early 2026. These are not yet comparable volumes, but the trajectory matters more than the current gap, particularly for businesses whose traffic depends on informational and research-driven queries.
Key facts to understand the shift in 2026:
- Google's global search share sits at roughly 89-91%, down from a peak of approximately 93% in 2021
- ChatGPT surpassed 1 billion daily queries as of late 2024, making it a direct competitor for informational search
- Perplexity AI's model combines cited sources with a synthesised answer, appealing strongly to research-heavy users
- Informational queries ("how", "what", "why") are the most vulnerable query type for Google to lose
- Google launched AI Overviews in May 2024 as a direct defensive response to LLM competition
Not all query types are equally at risk. Google retains dominant share in local, transactional and navigational search. Someone searching for a Cape Town restaurant, buying a product or navigating to a brand's website is still overwhelmingly using Google. The erosion is concentrated in the informational layer, where users prefer a synthesised, direct answer over a ranked list of blue links. This distinction is critical for any SEO strategy built in 2026.
Web SEM advises clients to optimise for both environments simultaneously, because the underlying strategies overlap more than they diverge. Structured content, clear entity signals and citability all improve performance in Google's traditional rankings and increase the likelihood of being cited by ChatGPT, Perplexity and Google's own AI Overviews. A results-driven approach means not choosing between Google and LLMs but building content that earns visibility in both. That is the practical reality of search in 2026, and it is where a data-driven SEO strategy delivers the most durable return.