Can a Fake Brand Win in AI Search? New Experiment Says Yes

Why It Matters

The study shows marketers can quickly claim AI‑search real estate for brand‑specific queries, yet competing for generic topics still requires traditional authority signals. Understanding each AI engine’s behavior is crucial for allocating content resources and protecting brand reputation in emerging LLM‑driven SERPs.

Key Takeaways

• Branded queries generated 96% of AI visibility for the fake brand
• Perplexity surfaced new pages in 1‑3 days but cited supporting domains
• Deep guides and review articles earned the most AI citations per page
• Thin, high‑volume pages yielded more total citations than a hub cluster
• Google AI Mode kept brand at #1 for 90% of branded prompts

Pulse Analysis

Artificial intelligence is reshaping how users retrieve information, turning traditional SEO tactics into a new set of signals that LLMs prioritize. While classic search engines still weigh backlinks and domain authority, generative models lean heavily on explicit brand mentions and structured, comprehensive content. This creates a dual‑track approach: brands must secure their narrative through dedicated "about" and "complete guide" pages that answer unique, brand‑centric questions, while also building depth in high‑value formats such as deep guides, reviews, and comparison pieces to attract citations across multiple AI platforms.

The experiment also highlighted stark differences among AI engines, underscoring the need for platform‑specific strategies. Perplexity’s rapid indexing rewarded fresh, high‑volume pages, yet its citation patterns favored ancillary domains, diluting brand focus. In contrast, Google’s AI Mode demonstrated remarkable stability, consistently surfacing the primary brand site for branded prompts, making it a reliable channel for brand control. ChatGPT showed a slower ramp‑up but eventually favored unique claims and comparison content, suggesting that sustained publishing can improve visibility over time. Gemini’s inconsistent performance and frequent misidentifications serve as a cautionary tale, reminding marketers that not all LLMs treat brand signals equally.

From a content planning perspective, the data suggests a hybrid model: invest in a few high‑quality, in‑depth assets to secure authoritative citations, while supplementing them with a broader set of shorter, topic‑aligned pages to increase overall retrieval odds. However, relying solely on topical clusters without sufficient citation signals proved ineffective, indicating that internal linking alone does not guarantee AI visibility. Brands should therefore balance depth with breadth, ensuring that core pages articulate brand identity clearly, while a steady stream of supplemental content keeps the brand top‑of‑mind across diverse AI answer engines.