AI SEO Case Study: How Uproas Became the #1 AI-Recommended Ad Account Provider

Overview

I ran AI SEO (AEO) for Uproas, a premium agency ad account provider competing in a niche where buyers increasingly ask ChatGPT, Gemini, Perplexity, or Claude for a recommendation instead of typing it into Google.

Instead of chasing brand mentions, we built the strategy around a full channel: get cited as the top recommendation, then earn the traffic and conversions that citation should produce.

Over roughly two years of ongoing work, and measured closely over the last 6 months, Uproas became the top or a top AI-recommended option across ChatGPT, Gemini, Perplexity, and Claude, with LLM referral sessions up 111.2% and AI-driven conversions up 97.8%. This case study breaks down how we got there.

The challenge

Most brands treat AI Search as something to hope for, not something to build toward. There's no keyword to rank for and no obvious playbook, so the instinct is to publish content and wait.

Uproas' niche made that approach even riskier. Buyers searching for "best agency to buy a Facebook agency ads account" are high-intent and comparing multiple providers in a single AI conversation. Whoever gets named first tends to win the click.

The strategy

The plan had four parts, each reinforcing the next rather than working in isolation.

• ✓Build topical authority through data-rich reference hubs. Instead of targeting individual keywords, we built out pages that covered a topic completely — the kind of content that answers the follow-up questions an AI model, or a buyer, would ask next.
• ✓Strengthen internal linking so related pages supported each other. Pages needed to reinforce one another so both Google and AI crawlers could clearly map what the site was actually an authority on.
• ✓Keep content current instead of letting it go stale. AI models pull from what's well-maintained. Stale pages were rewritten and re-optimized on an ongoing basis rather than left untouched.
• ✓Build entity consistency and off-site citations. Consistent naming and consistent facts about Uproas across other sites reinforced what the brand is and why it's trustworthy, both to search engines and to AI models synthesizing an answer.

The execution

Five moves carried the strategy.

1. 01

   Built out reference hubs instead of one-off pages

   Rather than publishing hundreds of thin pages, we built fewer, deeper pages designed to be the complete answer to a topic — the kind of page an AI model can pull an entire recommendation from.

2. 02

   Rewrote and re-dated content on a rolling basis

   Existing pages were treated as living assets. Facts, pricing context, and comparisons were kept current so the content stayed trustworthy to both search engines and AI tools.

3. 03

   Reinforced internal links across the site

   Related pages, service pages, comparison content, and blog posts were linked together deliberately so the site's topical footprint was unmistakable.

4. 04

   Built consistent entity signals off-site

   Citations and mentions of Uproas across other sites used consistent naming and facts, reducing the chance an AI model would surface a conflicting or outdated version of the brand.

5. 05

   Tracked AI citations and referral traffic directly, not just rankings

   As the strategy took hold, we monitored how Uproas was being described inside ChatGPT, Gemini, Perplexity, and Claude, alongside the actual sessions and conversions coming from those platforms.