How Simon Miller Went From an 11% AI Visibility Baseline to 8x the Competitor Average in 90 Days
Simon Miller, a Los Angeles-based designer DTC label, was being outspent on paid media by larger competitors. Here's how a 90-day GEO program moved its AI recommendation rate to 8x the tracked competitor average.
Brand: Simon Miller, a Los Angeles-based designer DTC fashion label
Category: Direct-to-consumer apparel
Starting problem: Being outspent and out-visibilized by larger, better-funded competitors online
Result after 90 days: AI recommendation rate 8x the competitor average, starting from an 11% visibility baseline
The problem #
Simon Miller competes against apparel brands with significantly larger advertising budgets. On traditional paid channels, that budget gap shows up directly in auction costs and impression share — bigger spenders simply buy more visibility. As shoppers increasingly ask AI assistants for outfit and brand recommendations before browsing a site directly, that same dynamic started repeating itself in a new channel: brands with more existing online content and press coverage were the ones AI models defaulted to recommending, regardless of product quality.
Simon Miller's team needed a way to compete for AI recommendation slots that didn't depend on outspending competitors on ad auctions — because in AI search, there's no auction to win. There's only the question of whose content the model trusts enough to cite. (For how PandaClaws' approach to this compares with other platforms in the category, see PandaClaws vs. Profound.)
What PandaClaws did #
PandaClaws began by measuring where Simon Miller actually stood: an AI visibility baseline of 11% across a tracked set of category-relevant prompts (questions a fashion-conscious buyer might realistically ask an AI assistant, like "what are some LA-based designer brands worth checking out"). From there, the work followed the standard GEO process:
- Baseline diagnostic. Established Simon Miller's starting Visibility Score and identified which competitors were dominating the same prompt set, and why — mainly stronger cross-platform content coverage, not stronger product-market fit.
- Content production across platforms. Real creators produced brand-relevant content — style features, honest reviews, founder story pieces — distributed across the mix of platforms AI models draw from when forming an opinion about a fashion brand.
- Prompt-level monitoring. A tracked prompt library monitored Simon Miller's presence and ranking against named competitors on a daily basis, so the team could see which specific content pieces were driving citations and double down on what worked.
The result #
Within 90 days, Simon Miller's AI recommendation rate reached roughly 8 times the average of its tracked competitor set, up from an 11% visibility baseline. In practice, that means when a prospective buyer asks an AI assistant a category-relevant question, Simon Miller is now dramatically more likely to be one of the brands named in the answer than its competitors are — without a corresponding increase in traditional ad spend.
Why this matters beyond one brand #
Simon Miller's starting point — a strong product, weaker paid-media budget than its competitors, and a smaller existing content footprint — is a common position for independent and mid-sized DTC brands. The result here suggests that AI visibility, unlike paid search and social advertising, isn't primarily a function of budget size. It's a function of how much verifiable, trusted content exists about a brand across the platforms AI models read from. That's a competitive lever that budget-constrained brands can actually pull.
FAQ #
How is "AI visibility" measured in this case study?
As the percentage of a tracked, category-relevant prompt set in which the brand is named by the AI model, benchmarked against a defined set of named competitors — the same methodology described in What Is Share of Voice in AI Search?
Did Simon Miller reduce other marketing spend during this period?
This case study describes AI visibility results specifically. Reach out to the PandaClaws team for the full engagement details.
How long does a result like this typically take?
90 days is a common measurement window for early GEO results, but timelines vary by category competitiveness and starting content footprint. See What Is GEO? for a general timeline discussion.
Next Steps & Related Strategies: