Documented Results

AI Visibility Is Winnable.
Here's the Proof.

Every case study here is based on a real AI Visibility Snapshot audit. Real businesses. Real before-and-after data. Sectors named, client names protected.

340+
AI Overview appearances
in one 90-day window
4x
Brand mention share increase
in zero-click AI responses
3
AI platforms tested per audit
(ChatGPT, Perplexity, Gemini)
Methodology

Each case study uses the Sticky Frog AI Visibility Snapshot, a structured 47-point audit across schema markup, llms.txt presence, content architecture, entity definition, and live AI citation testing across three platforms. Results are documented at point of audit and verified 30 days post-implementation where possible.

Case Studies
Case Study 01  /  Professional Services / Accountancy

A Regional Accountancy Firm With Strong Google Rankings and Zero AI Presence

Established firm (10+ years), ranking on page 1 for local search terms, active blog, professional website. Had invested significantly in traditional SEO. Not a single AI platform was recommending them when prospective clients asked "who are the best accountants in [region]?" or "can you recommend an accountancy firm for my small business?"

AI Visibility Snapshot Audit
Issues Found
6 of 6
Structural AI visibility issues
identified and prioritised
Before the Snapshot
  • ✗No llms.txt file: AI crawlers had no structured guidance on what content to prioritise
  • ✗No FAQ or HowTo schema: key service pages had zero structured data for AI consumption
  • ✗Keyword-optimised, not question-optimised: content answered no direct question a prospective client would ask an AI
  • ✗No entity definition page: AI systems could not confidently identify who the firm was, what they specialised in, or where they operated
  • ✗Zero AI citations: tested across ChatGPT, Perplexity, and Gemini for 12 relevant queries. Zero citations across all three.
  • ✗Features, not outcomes: all copy described services, not the results clients experience. AI systems cite outcome language, not feature lists.
What the Snapshot Delivered
  • ✓Prioritised fix list: 6 issues ranked by AI visibility impact, with exact implementation steps for each
  • ✓llms.txt file specification: custom-written and ready to deploy, structured to guide AI crawler behaviour
  • ✓FAQ schema template: for the firm's top 5 service pages, pre-written with AI-optimised question-answer pairs
  • ✓Entity page blueprint: structure and copy guidance for an About page that AI systems can use to identify and recommend the firm
  • ✓Content rewrite brief: for their three highest-priority pages, shifting from feature language to outcome language
  • ✓Query cluster map: 15 questions their ideal clients are asking AI platforms right now, with gap analysis showing where the firm had no answer on record

The Insight

"This firm had done everything right for Google and it had worked. But Google's algorithm and AI's retrieval logic are fundamentally different. Google rewards domain authority and keyword density. AI rewards entity clarity, question-format content, and structured data. This firm was invisible to AI not because of poor content but because their content was built for the wrong reader entirely."

Jason Morris, Sticky Frog

Case Study 02  /  B2B SaaS / Workforce Technology

A B2B SaaS Platform Generating 340+ AI Overview Appearances in 90 Days

A UK-based B2B SaaS business with 3+ years of SEO investment, solid domain authority, and a content team producing articles consistently. Buyers in their space increasingly research via AI before requesting demos. The problem: when those buyers asked "what's the best workforce management software for mid-market businesses?" the platform never appeared. Their sales team kept hearing the same thing from new prospects: "We searched on ChatGPT first and only found you through LinkedIn."

Agency Client
90-Day Result
340+
AI Overview appearances
across tracked query clusters
The Challenge
  • ✗Content written for Google, not AI: three years of SEO articles targeted keywords but contained no direct answers to the comparative questions buyers were asking AI platforms
  • ✗No AI Overview optimisation: the site had never been evaluated for AI Overview eligibility. No FAQ schema, no HowTo markup, no concise answer blocks AI systems could extract
  • ✗Weak entity definition: despite a detailed product page, AI systems had no clear signal about what problem the platform solved, for whom, or how it compared to named alternatives
  • ✗Query fan-out blind spot: when AI decomposes a buyer's question into sub-queries, the site had no content mapped to those sub-queries. No comparison pages, no use-case pages, no "who is this for" content
  • ✗Zero AI citation at audit: tested across 15 high-intent query clusters. Appeared in zero AI-generated answers across all three platforms.
What Changed and the Results
  • ✓Query cluster content build: mapped 15 buyer query clusters and created dedicated answer pages for each, structured with direct-answer headings, FAQ schema, and outcome-first copy. AI systems began retrieving these pages within weeks.
  • ✓AI Overview schema rollout: FAQ and HowTo schema deployed across 12 core pages; confirmed eligible for AI Overview inclusion in Google Search Console within 30 days
  • ✓Entity disambiguation page: dedicated page defining the product category, target customer, key differentiators, and use cases in AI-readable format; cited by Perplexity in the first week after indexing
  • ✓340+ AI Overview appearances: tracked across 15 query clusters over a 90-day window, with appearances confirmed in ChatGPT, Perplexity, and Google AI Overviews
  • ✓12 of 15 query clusters won: cited or recommended in AI-generated answers for 12 of the 15 tracked buyer queries, up from 0 of 15 at audit
  • ✓Demo pipeline impact: inbound demo requests attributed to AI-referred traffic increased quarter-on-quarter, with prospects citing AI platform recommendations in discovery calls
340+
AI Overview appearances
in 90 days
12/15
Target query clusters
won in AI results
0 to 3
AI platforms now citing
the platform by name

The Insight

"B2B SaaS buyers are using AI as a pre-qualification tool. They ask ChatGPT or Perplexity to shortlist their options before they ever visit a vendor website. If your product is not in the AI's shortlist, you do not get evaluated. The brands winning this are not the ones with the biggest marketing budgets. They are the ones whose content is structured to answer the exact questions buyers ask at that research stage. Query fan-out is the mechanic: AI decomposes a buyer's question into sub-queries before synthesising a response. You have to map your content to those sub-queries, not the surface-level keyword."

Jason Morris, Sticky Frog

Case Study 03  /  Health & Wellbeing

How a Health and Wellbeing Brand Won AI Visibility in a Zero-Click Landscape

A direct-to-consumer health and wellbeing brand with a natural supplements range, operating in a category where AI search has fundamentally changed the game. When someone asks "what are the best supplements for [condition]?" they rarely click through to a product page. They get an AI-generated answer that either mentions your brand or does not. This brand had strong organic rankings, an engaged audience, and a credible product range but no presence in AI-generated wellness answers and no strategy to earn one.

Agency Client
Brand Presence
4x
Brand mention share in
zero-click AI responses
The Challenge
  • ✗Zero-click reality ignored: the business measured success by organic clicks and sessions. In health and wellbeing, AI now answers most intent queries without a click. The brand had no metric for AI brand mentions and no strategy to earn them.
  • ✗No expert author attribution: product pages and editorial content had no named author, no credentials, no E-E-A-T signals. AI systems prioritise content with verifiable human expertise, particularly in YMYL health categories.
  • ✗No third-party citation profile: the brand existed almost exclusively on its own domain. No editorial mentions in health publications, no named expert affiliated with the brand that AI could reference as an external source.
  • ✗Generic, non-citable copy: product descriptions were sales-led, not information-led. AI systems retrieve and cite information. They do not cite marketing copy.
  • ✗Brand not defined as an entity: AI systems had no structured information about what the brand stood for, who it was for, or what made it distinct. Without entity definition it was interchangeable with dozens of competitors.
The Authority-Building Programme
  • ✓Expert author pages built: a qualified nutritionist affiliated with the brand was given a dedicated author page covering credentials, professional background, and specialisms. AI systems in health categories now had a citable human expert to attribute content to.
  • ✓Informational content restructure: the top 10 product categories were given companion editorial pages structured as direct-answer Q&A content, each with cited sources and FAQ schema
  • ✓Third-party citation campaign: editorial placements secured in three relevant health and wellness publications, each linking to and naming the brand in context, giving AI systems a corroborated entity signal beyond the brand's own domain
  • ✓Brand entity page: a structured brand definition page deployed covering what the brand is, who it serves, and what makes it distinct. AI systems now had a canonical reference point for the brand as a named entity.
  • ✓4x brand mention share: in tracked zero-click AI responses across 20 wellness query clusters, the brand's mention rate increased from 1 in 20 queries to 4 in 20 queries over a 6-month window
  • ✓Cited by name on all three platforms: the brand began appearing by name in ChatGPT, Perplexity, and Gemini responses for relevant health queries, with the expert author cited as attributed source in Perplexity's cited-source format
4x
Brand mention share
in AI zero-click responses
3/3
AI platforms now citing
the brand by name
6 mo
Authority programme
timeframe

The Insight

"Health and wellbeing is the hardest vertical for AI visibility and the most important one to get right. AI platforms apply YMYL standards to health content, which means the bar for citation is higher. You cannot just add schema and hope. You need a verifiable human expert, third-party corroboration, and content that is genuinely informational rather than sales-led. The brands who invest in this properly do not just earn AI citations. They build the kind of authority profile that drives sustainable trust across every channel. In a zero-click world, the goal is not the visit. It is the mention. And the mention compounds."

Jason Morris, Sticky Frog

What the Snapshot Audits

Every AI Visibility Snapshot covers 47 structural checks across six categories, the same methodology used in every case study above.

🤖

Live AI Citation Testing

Your site tested across ChatGPT, Perplexity, and Gemini for 12+ relevant queries. We record exactly what AI systems say about you and why.

🏗️

Schema & Structured Data

FAQ, HowTo, Article, Organization, and Person schemas checked for presence, accuracy, and AI-retrieval readiness.

📄

llms.txt & Crawler Access

AI crawlers need instruction. We check whether your llms.txt exists, is structured correctly, and guides AI to your best content.

✍️

Content Architecture

Is your content written for AI retrieval? We check question-format structure, outcome language, entity definition, and answer density.

🎯

Entity Definition

Can AI systems confidently identify who you are, what you do, and where you operate? Entity clarity is the single biggest lever for AI citation.

📊

Priority Action Plan

Every audit ends with a ranked implementation plan. Highest AI visibility impact first, with exact steps your team can execute immediately.

Your Site Has the Same Issues.
Find Out Which Ones.

The AI Visibility Snapshot takes 5 minutes to book and delivers a complete picture of your AI search presence, with a prioritised fix list your team can action immediately.

Get Your Free Snapshot See How It Works

Get the Research Before Anyone Else

The Human Algorithm newsletter delivers AI visibility findings from real audits. Patterns, data, and tactics drawn directly from live Snapshot work. Published monthly. No fluff.

Join practitioners getting real AI search data. Unsubscribe any time.

Scroll to Top

Frequently Asked Questions

Common questions about AI search, AEO, and how Sticky Frog helps B2B businesses get cited by AI engines.

What is AEO (Answer Engine Optimisation)?

AEO stands for Answer Engine Optimisation. It is the practice of structuring your website content, entity data, and online presence so that AI search engines like ChatGPT, Perplexity, and Google AI Overviews cite your business in their generated answers. Unlike traditional SEO, which targets click-through traffic, AEO targets citation: being the source an AI engine recommends when someone asks a relevant question.

Why does AI search visibility matter for B2B businesses?

B2B buyers increasingly use AI tools like ChatGPT and Perplexity to generate vendor shortlists before making contact. If your business is not cited by these AI engines, you are invisible to these buyers at the most critical point in their decision-making process. AI shortlisting makes AI search visibility a strategic priority for any B2B business.

What is the difference between SEO, AEO, and GEO?

SEO focuses on ranking in traditional Google search results. AEO (Answer Engine Optimisation) focuses on being cited in AI-generated answers on ChatGPT and Perplexity. GEO (Generative Engine Optimisation) focuses on appearing in outputs of generative AI tools. Sticky Frog specialises in AEO for B2B businesses and professional services.

What is an llms.txt file and does my website need one?

An llms.txt file is a plain-text file at the root of your domain that tells AI language model crawlers what content to index, trust, and cite. It is the AI equivalent of robots.txt. Most business websites do not yet have one, making it a meaningful competitive advantage in AI search visibility.

How long does it take to see results from AEO?

AI search visibility improvements can begin within 4 to 8 weeks for technical fixes like schema markup and llms.txt. Content-driven citation builds over 3 to 6 months. The AI Visibility Accelerator is a minimum 6-month engagement delivering results across ChatGPT, Perplexity, Google AI Overviews, YouTube, and Reddit.