Unusual AI
Unusual AI is positioning as a seed horizontal AI infrastructure play, building foundational capabilities around continuous-learning flywheels.
With foundation models commoditizing, Unusual AI's focus on domain-specific data creates potential for durable competitive advantage. First-mover advantage in data accumulation becomes increasingly valuable as the AI stack matures.
Unusual AI is an IT and Internet firm offering AI that personalise website according to visitors for more conversions.
Unusual AI's unique advantage is its direct surveying and analysis of LLMs' opinions about brands, combined with targeted content engineering to change those opinions and drive qualified leads.
Continuous-learning Flywheels
Unusual AI implements a feedback loop where they survey AI models to understand current perceptions, analyze the results, and then generate targeted content to influence future model outputs. This iterative process leverages continuous feedback from AI model outputs to refine and improve how brands are represented in AI responses.
Winner-take-most dynamics in categories where well-executed. Defensibility against well-funded competitors.
Vertical Data Moats
Unusual AI focuses on creating and leveraging proprietary, brand-specific datasets and content to influence AI model outputs. Their approach is tailored to specific industries and brands, creating a competitive advantage via unique, domain-specific data and insights.
Unlocks AI applications in regulated industries where generic models fail. Creates acquisition targets for incumbents.
RAG (Retrieval-Augmented Generation)
There are indications that Unusual AI is aware of and optimizing for the types of content that retrieval-augmented generation systems prefer, such as long-form documentation and API docs, to influence how AI models generate responses. However, there is no direct mention of retrieval pipelines or vector search.
Accelerates enterprise AI adoption by providing audit trails and source attribution.
Unusual AI builds on ChatGPT, Copilot, LLMs. The technical approach emphasizes prompt engineering.
Unusual AI operates in a competitive landscape that includes Market Brew, Yext, BrightEdge.
Differentiation: Unusual AI focuses on shaping how large language models (LLMs) like ChatGPT talk about brands, treating AI models as influencers and a new audience, rather than optimizing for search engine rankings. Market Brew is primarily SEO-focused and models search engine algorithms rather than LLMs.
Differentiation: Unusual AI targets AI models (LLMs) as the new channel for brand discovery and recommendation, whereas Yext is focused on structured data for search engines and directories, not conversational AI or LLMs.
Differentiation: Unusual AI's approach is not about keyword ranking or prompt volume, but about probing and shaping the opinions and recommendations of AI models, which personalize responses and act as 'influencers' rather than search engines.
Unusual AI treats AI models as 'influencers' rather than search engines, suggesting a paradigm shift in how brands approach AI visibility and reputation management. This reframing is technically interesting because it implies the need for continuous, dynamic interaction with LLMs, rather than static SEO-like optimization.
The platform appears to use 'brain probe' prompts—custom batteries of queries designed to elicit specific opinions, biases, and source preferences from different AI models. This is a novel approach to model auditing and brand perception analysis, moving beyond simple prompt engineering to systematic model interrogation.
Unusual claims to reverse engineer the content and data landscape needed to change AI model perceptions, implying a feedback loop between model output analysis and targeted content creation. This suggests a semi-automated pipeline for discovering gaps in model knowledge and rapidly generating corrective content.
The system's ability to survey multiple AI models and analyze their 'opinions' about brands hints at a meta-model layer or orchestration framework that can compare, contrast, and synthesize outputs from different LLMs. This is technically non-trivial, especially at scale and across diverse model architectures.
There is an emphasis on long-form, high-clarity content (such as API documentation) as a preferred source for LLMs, which is a subtle but important insight for brands seeking to influence AI-driven recommendations. This reflects a deeper understanding of LLM retrieval and synthesis behavior.
The product appears to be primarily a layer on top of existing LLMs (e.g., OpenAI, Anthropic), focused on prompt engineering and 'surveying' AI models. There is no clear evidence of proprietary models, infrastructure, or deep technical innovation beyond orchestrating prompts and aggregating responses.
The core offering—shaping how AI models talk about your brand—could be implemented as a feature within broader marketing or SEO tools, or even by LLM providers themselves. The scope appears narrow and could be subsumed by incumbents.
There is no clear data, network, or technical moat. The approach relies on public LLM APIs and prompt engineering, which are easily replicable. No evidence of unique datasets, vertical data moats, or defensible IP.
If Unusual AI achieves its technical roadmap, it could become foundational infrastructure for the next generation of AI applications. Success here would accelerate the timeline for downstream companies to build reliable, production-grade AI products. Failure or pivot would signal continued fragmentation in the AI tooling landscape.
Source Evidence(10 quotes)
"Unusual is more than an AEO/GEO tool. It’s like teaching ChatGPT how to sell your product for you."
"Unusual’s approaching GEO in a new way, and it’s been incredibly helpful in quickly updating how LLMs learn and cite us."
"Unusual’s content is Chatgpt’s go-to reference for us! We now have much more control on how LLMs represent us."
"We doubled the number of leads from Copilot within two months of going live with Unusual"
"Unusual surveys AI models to discover how they see your brand, and then suggests actions to shape their view."
"Build a battery of “brain probe” prompts specific to your business, market, and competitors in order to reveal AI’s thought patterns, sources, biases, etc."