Walmart leads 100 most-cited brands in ChatGPT, Ahrefs finds

Walmart appears in 95,605 of the 5.5 million US ChatGPT answers Ahrefs sampled. Target, in second place, trails by nearly half. That gap is the story. ChatGPT's brand mentions are not a popularity contest with marginal differences at the top; they compress hard around a handful of category leaders and then fall off a cliff.

Ahrefs's method is blunt and useful: count brand name occurrences across a large, representative pool of real consumer prompts, then rank. No sentiment analysis, no weighting for favourability, just raw frequency of being named. Stores take six of the top ten slots. Carmakers account for 15 of the 100. Retail and automotive, in other words, dominate a list that is supposedly about the entire US brand economy. That is not because Americans ask ChatGPT more about toasters than banks. It is because ChatGPT's training data and retrieval patterns favour categories with dense, structured, frequently updated public information: product catalogues, reviews, comparison content, recall notices, dealer inventories. Walmart and Target generate an enormous footprint of exactly that kind of content. A regional insurer or a mid-sized asset manager does not, and so it barely registers even when users are actively asking about financial products.

The category tax nobody is pricing in

This has a direct and under-examined consequence for financial services, multilaterals, and industrial groups: your category is structurally disadvantaged in LLM answers before you have done anything wrong. A bank's products are commoditised by regulation and hedged by compliance language; the public web rarely contains the dense, comparison-friendly content that retail generates by default (spec sheets, star ratings, "X vs Y" pages). The same is true for a reinsurer, a standards body, or a UN agency: nobody is writing 50,000 consumer-facing comparison pages about ISO's technical committees or CGAP's microfinance guidance, even though the underlying subject matter is more consequential to GDP than another listicle about air fryers.

The result is a visibility floor that retail blows through and everyone else has to fight to reach. Ahrefs's data shows what the ceiling looks like when a brand does things right: Walmart and Target do not win because they advertise more, they win because their sites document an enormous number of answerable questions (price, availability, return policy, comparison to competitors) in a form models can extract cleanly. That is a content-architecture advantage, not a brand-awareness one. It is replicable by any institution willing to build out the same kind of structured, question-shaped content for its own domain: fee schedules, risk disclosures, standard specifications, programme eligibility criteria. Most financial and multilateral institutions still publish this material as PDFs, buried three clicks deep, written for regulators rather than retrieval systems.

Why the carmaker cluster matters more than it looks

The automotive presence (15 of 100) is worth a second look, because cars share a structural trait with financial products that retail goods do not: high consideration, long research cycles, comparison-heavy decision-making. If carmakers can claim 15 slots on raw mention volume, it confirms that LLMs do surface considered-purchase categories when the content exists in the right shape; Toyota and Ford are not winning by being cheap impulse buys, they are winning because dealer sites, review aggregators, and manufacturer spec pages between them generate the kind of structured comparative content a model can lift wholesale into an answer. Insurance, wealth management, and industrial procurement are no less "considered" than a car purchase. They are simply less documented in machine-readable form.