How modern AI systems retrieve information, generate answers, and influence which brands get recommended.
Consumers are increasingly asking AI what to buy before they search, browse, or compare.
The brands surfaced in those answers shape which products buyers consider first.
For ecommerce brands, understanding how recommendation systems work is becoming part of understanding how products get discovered.
Plain-language explanations of how modern AI systems work — no technical background required.
A plain-language overview of how modern AI systems work — and why they behave differently from traditional software.
How large language models like ChatGPT, Claude, and Gemini generate responses — and why they power most modern AI experiences.
A high-level overview of how training data, retrieval, and reasoning work together to produce AI-generated responses.
Understanding the sources behind AI answers — and why it matters for brand visibility.
Why training data influences what an AI system already understands about brands, products, and concepts — before a question is ever asked.
How modern AI systems use training data, retrieval systems, browsing, and external sources to generate answers — and why different source types affect brand visibility differently.
A deeper explanation of the difference between learned knowledge (baked into the model) and retrieved information (looked up at query time) — and why the distinction matters.
How ChatGPT may use training data, browsing, memory, and retrieved context depending on the situation — and how that shapes which brands it surfaces.
How the major AI systems compare in their approach to source use, retrieval behaviour, and brand recommendation patterns — and what those differences mean for ecommerce visibility.
The bridge between AI education and ecommerce strategy — how recommendation systems affect which brands buyers see.
A practical introduction to AI visibility — why recommendation systems are becoming part of modern product discovery and what it means for ecommerce brands.
Why an AI system's understanding of a brand may be incomplete, outdated, or inconsistent — and how that directly affects recommendation inclusion.
How architecture, training data, and retrieval design lead to meaningfully different recommendation outputs across platforms.
What tends to drive shifts in which brands an AI system surfaces — and why recommendation visibility is dynamic rather than fixed.
Why ranking in search results and appearing in AI recommendations are often different outcomes — and why recommendation visibility is becoming its own category of discovery.
The concepts in this library are regularly observed inside PickMeLabs experiments.
Illustrates how recommendation patterns concentrate around a small number of brands across different query types.
Shows how buying context and geography shift which brands appear across the same category queries.
Demonstrates how well-known celebrity-backed brands can remain absent from AI recommendation environments.
We're building a growing library of educational resources about AI-powered discovery. If there's a topic you'd like explained, let us know.
Selected suggestions may become future AI Basics articles, research notes, or experiments.
The PickMeLabs Visibility Experiment helps ecommerce teams understand where their brand appears, where competitors are being surfaced, and what patterns may be influencing recommendation inclusion.
Designed for brands that want a deeper understanding of AI visibility, recommendation behaviour, and competitive positioning.
The proprietary frameworks used throughout PickMeLabs research and experiments.
Long-form analysis explaining recurring patterns observed across AI recommendation environments.
How PickMeLabs designs experiments, evaluates outputs, and studies recommendation behaviour.