Client Zero · Living Case Study

AISearch Global: Client Zero — From 19 to 91, and What AI Still Doesn't Know

Before selling AEO to anyone, we had to prove it on ourselves. Client Zero is AISearch Global's own brand — tracked from day one, 20 May 2026, with every signal, every fix, and every score run exactly the way a paying client's audit runs.

By Viveka Mohan Das · Published 21 Jun 2026 · Next update ~21 Sep 2026

 Visibility Audit Dashboard

The numbers in this article come from the interactive AI Visibility Audit Dashboard — 10 audit instruments with drill-down panels for every score. Click any signal card to see the full data behind it.

Open the Visibility Audit Dashboard

Starting From Zero

AISearch Global launched on 20 May 2026. No content backlog. No prior schema. No entity history for AI engines to draw on. A clean baseline by design. If we can't move our own brand's AI visibility using the exact AEO Traction Stack we sell, we have no business selling it to anyone else — so we ran the audit on ourselves first.

Day one score: 19/100 on our own AEO Score Calculator. That's not a humble-brag number — it's a genuinely weak starting position, the same range most Australian SMB sites land in before any AEO work begins.

What We Actually Measure

The AEO Score Calculator checks 20 signals — 17 structural AEO signals plus 3 signals inspired by the GEO (Generative Engine Optimisation) research. In plain language, they group into the four layers of the AEO Traction Stack:

Entity Clarity

Can an AI engine state, without guessing, who you are, what you do, and where you operate? This is usually the single biggest gap on a new site and accounts for the largest share of early score movement.

Schema Markup

Structured data (JSON-LD) that lets AI systems extract facts directly — your services, your location, your credentials — instead of inferring them from page text.

Answer-Format Content

Content written the way AI engines lift answers: direct questions paired with direct answers, rather than long, undifferentiated prose.

Citation Consistency

Whether your business name, services, and location are described the same way everywhere AI engines can see them — your site, directories, social profiles. Inconsistency erodes trust signals even when each individual mention is accurate.

Layers one and two — entity clarity and schema — account for roughly 80% of the improvement most brands see. That's exactly where our own six weeks of work went first.

Methodology

The AI Brand Perception Audit scores in this article were collected using a consistent protocol applied identically across all three platforms. For reproducibility:

Every quarterly update will use the same protocol against the same prompt set, so movement between periods reflects actual change rather than methodological drift.

The Climb: 19 → 91

Structural AEO score — implementation milestones

20 May 2026 → 20 Jun 2026

Launch (20 May) 19/100
Site + Schema (18 May) 45/100
Content Wave 1 (30 May) 68/100
FAQ + Entity (7 Jun) 79/100
Full Stack (20 Jun) 91/100

+72 points in six weeks, built on weekends and evenings around a full-time job. No agency team, no paid sprint. The score jumped fastest in the first two weeks (entity clarity and schema alone took it from 19 to 45) — exactly as the AEO Traction Stack predicts. Here's exactly what that work looked like, including the things that went wrong:

None of these were dramatic. They're the ordinary friction of building a structured-data site — and exactly the category of fault an AI Visibility Audit catches before it quietly costs a client citations.

The Score That Didn't Move as Fast

A structural score of 91 tells us the page is machine-readable. It doesn't tell us what AI engines currently know. The second instrument in our methodology — the AI Brand Perception Audit — asks ChatGPT, Perplexity, and Gemini directly, rather than scoring what's on the page and assuming they've read it.

Why these scores matter more than they look

AISearch Global went live on 20 May 2026. The base training data for these models largely predates AISearch Global's launch — although platforms may supplement responses using live retrieval, the foundational knowledge each model draws on was formed before we existed.

Yet all three platforms found, identified, and scored us. Correctly. A brand that didn't exist when the AI was trained is still appearing in AI answers weeks after launch. That is not a given — most new brands don't register at all. Structural AEO is the most plausible explanation — making the site legible to AI crawlers from day one is consistent with the impact expected from strong structural implementation.

ChatGPT

Base training predates our launch

Supplements via live web access — finds us despite no training data

Gemini

Base training predates our launch

Scores us highest (41) — richer Australian tech context in training data

Perplexity

No cutoff — live web

Real-time indexing; first to reflect structural changes

The scores of 36, 30, and 41 are not low despite the work done — they are high given that we launched after every model's training cutoff. The structural baseline we built is already being read. The next phase is citation volume: giving the models more to say about us.

Why recognition is low · the timing gap

AI training datasets
knowledge frozen here
AI's blind spot
no new data ingested
22 May 2026
Business & online presence live
Graded
20 Jun 2026
This is the headline: the AI models were trained before we existed — yet even with zero training-data history, all three engines still found and rated the brand. That is a strong starting position, not a weak one.

ChatGPT

36/100

Partial recognition

Perplexity

30/100

Limited recognition

Gemini

41/100

Partial recognition

Score breakdown — 5 dimensions behind 36 / 30 / 41

Dimension ChatGPT Perplexity Gemini Max
Brand Recognition 2 1 2 /20
Market Score 6 6 6 /10
Presence Quality 5 3 5 /20
Brand Sentiment 22 18 27 /40
Share of Voice 1 2 1 /10
Total 36 30 41 /100

Brand Sentiment (weighted /40) drives most of the platform gap — Gemini's 27 vs ChatGPT's 22 vs Perplexity's 18. Market Score is tied at 6/10 across all three: category relevance confirmed. Brand Recognition in single digits on every platform — the primary Q3 target.

Interpretation note: These scores should be read as directional benchmarks rather than laboratory-grade measurements. They reflect a structured assessment at a single point in time using consistent prompts and scoring criteria — useful for tracking movement, not for precise comparison between platforms.

A second, related score — GEO AEO Visibility, measuring whether AI engines anchor the brand as Australian and Sydney-based rather than generic — sits at 45/100 ("Emerging Local Authority"). The four sub-scores tell a more granular story:

GEO Presence 60/100
Entity Localisation 49/100
Conversion Readiness 36/100
Local Intent Coverage 31/100 ⚠

Per-platform GEO scores: Gemini 56/100 · Perplexity 43/100 · ChatGPT 36/100. Gemini scores highest on local GEO despite having the oldest training cutoff — because its pre-launch training data already contains a richer anchor of Australian tech ecosystem context. Perplexity beats ChatGPT because its real-time access picks up locally-framed content as it's published. Local Intent Coverage at 31/100 is the clearest action item: more locally-anchored content, structured location signals, and directory citations pointing at Sydney.

Why the gap is real, not a contradiction: 91/100 measures what we control directly — the page itself. 36/30/41 measures what AI models currently know, and those models don't re-crawl and re-learn the web in real time. There's a lag between a page becoming machine-readable and a model actually citing it, shaped by training cutoffs and how often each platform refreshes its citation sources. Fixing the page is necessary but not sufficient — closing that lag is ongoing work, not a one-off fix.

The perception gap — why 91 and 36/30/41 aren't a contradiction

The gap between our 91 structural score and 36/30/41 AI brand perception is the normal relationship between two independent clocks: the structural clock, which we control and moved 72 points in six weeks, and the perception clock, which the AI platforms control on their own training and retrieval cycles. Fixing the page is necessary; it isn't sufficient.

This is the two-clock model, and it's the framework concept most agencies can't explain — because most only sell structural fixes and never run a perception audit. Full breakdown: The Two-Clock Model.

What AI Already Knows

We asked ChatGPT, Perplexity, and Gemini directly: who is AISearch Global, and what do they do? The answers are a useful diagnostic in their own right — not just three numbers.

Getting it right

All three platforms correctly identify AISearch Global as a Sydney-based AI visibility / AEO specialist — not a generalist marketing or SEO agency. Founder name, service category, and Australian location come through consistently wherever the brand is recognised at all.

Still missing

None of the three platforms can cite a specific case study, client result, or independent third-party mention yet. That's the honest shape of a brand-new entity: strong structural signals, no external citation history. Closing that gap is precisely what the Citation Consistency layer of the AEO Traction Stack is built for.

That's the practical meaning behind 36/30/41: the models know what we say about ourselves. They don't yet know what anyone else says about us. This article — and the dashboard behind it — is itself one of the first citable data points working to close that gap.

What AI Thinks About the Brand — Archetype & Sentiment

Beyond raw scores, each platform was asked to characterise the brand's positioning. All three consistently characterised AISearch Global as innovative — drawing on different data sources and arriving at the same directional framing.

Brand Archetype: Innovator

ChatGPT (84% confidence), Perplexity (60%), and Gemini (35%) all independently classify AISearch Global as an Innovator in a niche market with a 6/10 competitive intensity score. Confidence varies because data volume varies — but the convergence of identity across three different platforms with different data sources is itself a strong structural signal.

Sentiment: Gemini leads, others building

Gemini scores sentiment at 68/100 — likely drawing on positive industry framing from late 2024. ChatGPT sits at 54/100 and Perplexity at 45/100. All three show low polarisation scores (24–35/100), which is ideal: AI platforms tend to be cautious about brands with divisive framing. Low polarisation means the brand reads as credible and non-controversial.

Share of Voice Against Established Competitors

A newly launched brand appearing in AI responses about the category at all is notable. Against established players, AISearch Global holds 8% SOV on ChatGPT (vs WebFX at 18%), 18.5% on Perplexity (vs ThatWare at 22%), and 8.5% on Gemini (vs Search Engine Land at 22%). Perplexity's real-time indexing gives the brand its highest share — 142 mentions vs 42 on ChatGPT and 1,240 on Gemini, where established players have had years of training data advantage.

Why Perplexity SOV is highest at launch: Perplexity reads the live web continuously rather than relying on a training cutoff. New citations appear almost immediately in its answers. This makes citation-building work visible faster on Perplexity than on ChatGPT or Gemini — and makes it the first platform to reflect structural improvements.

The Machine Is Already Watching — AI Crawler Activity

Six AI crawler bots actively indexed the site in the 30-day audit window: PerplexityBot, ChatGPT-User, Claude-SearchBot, ClaudeBot, OAI-SearchBot, and GPTBot. AI crawlers accounted for 43 of 108 total crawler requests during this measurement period — 40% of all crawler traffic — which is significantly above average for a one-month-old domain. Crawler volume increased +2,600% over the period.

What this signals — and why it matters: A domain that went live on 20 May 2026 is already being crawled by every major AI bot, with a +2,600% crawl growth rate, before it has had time to accumulate any citation history. The gap between a high structural score and lower AI perception scores is not because the platforms aren't looking. They are actively reading the site. The lag exists because AI training data has fixed cutoffs and citation volume is still thin — not because the work isn't landing. For a brand that postdates every model's training corpus, appearing in AI answers at all within weeks of launch is the result to focus on, not the scores. The machine is already watching. The work now is giving it more to say.

One technical flag: the Cloudflare cache rate sits at 33% against a target of 70–85%. AI crawlers hit origin servers more often than they should — slower and less reliable. A cache header fix is on the roadmap and resolves in days, not weeks.

How AI Engines Describe the Brand

Beyond scores, each platform was asked to characterise AISearch Global in its own words. These themes are a diagnostic of what's landing — and what needs building before the brand moves from category-correct to specifically-cited.

ChatGPT

  • "AI-driven search visibility and AEO expertise"
  • "Helping brands adapt to evolving answer engines"

Category-correct. Brand-specific detail not yet there — no named results or client outcomes attributed.

Perplexity

  • "AI-Driven Search Visibility Strategy"
  • "Answer Engine Optimization Expertise"

Real-time access, but mirrors ChatGPT closely — limited citation depth to draw richer themes from yet.

Gemini

  • "Pioneering AEO in the Australian tech landscape"
  • "Navigating the shift from traditional SEO to AI search"
  • "Empowering global brands to dominate generative AI results"
  • "Leading the transition to conversational search"

Four distinct themes — richer and more narrative. Correctly positions the brand as Australian and forward-leaning. Oldest cutoff, but the strongest descriptive framing of any platform.

Market trajectory — all three platforms agree: AISearch Global is well-positioned as AI search adoption accelerates. The window for establishing first-mover authority in Australia is open now — before the market saturates with generalist agencies pivoting to AEO.

This Will Keep Changing

Every number on this page is a snapshot. AI platforms update how they crawl, weight, and cite sources on their own schedules — no public changelog, no warnings. A score can shift without the site changing at all. That's why we're running this audit on a fixed cadence, not publishing once and calling it done.

Audit & Update Timeline

20 May 2026
Baseline audit
Launch. AEO Score Calculator: 19/100. AI Brand Perception Audit: ChatGPT 36, Perplexity 30, Gemini 41.
21 Jun 2026
This article
Structural score reaches 91/100, published today as the baseline every future audit will be measured against.
~21 Sep 2026
3-month update
Re-audit and movement vs. baseline across ChatGPT, Perplexity, and Gemini — the next entry here.
~21 Dec 2026
6-month update
Second re-audit. A trend line across two checkpoints, not just a single one-off snapshot in isolation.
~21 Jun 2027
12-month update
One full year of tracked movement — the clearest proof point for sustained AEO work paying off.

Why This Matters for You

Most businesses have neither number. No structural AEO score because no one has run one, and no AI Brand Perception score because the questions have never been put directly to ChatGPT, Perplexity, or Gemini. That means they are making decisions about a gap they cannot see — assuming either that AI is already finding them, or that it is not something worth measuring yet.

The structural gap — the 19/100 range most unoptimised sites start at — is the faster fix. Six weeks of focused work moved ours 72 points. The perception gap is a different clock: AI models do not re-learn the web in real time, so even after a page is fully machine-readable, there is a lag before that model's answers reflect it. The earlier the structural work is done, the earlier that lag starts counting down. Waiting does not pause the gap — it extends it.

The businesses running both instruments in mid-2026 will have a six- to twelve-month head start on the ones who wait until AI search feels urgent. That is the window this article is written into — and it is still open.

Frequently asked questions

How do I know if AI systems are currently recommending my business?
The simplest way is to ask directly. Open ChatGPT, Perplexity, or Gemini and type queries your customers would ask — for example, "best AEO consultant in Sydney" or "who helps businesses get found by AI." If your business does not appear, you have a perception gap. The free AEO Score Calculator grades the structural signals that determine whether you appear. An AI Visibility Audit goes deeper, testing your actual citation rate across platforms — the same two-instrument method used in this case study.
What's the difference between the free AEO Score Calculator and the AI Visibility Audit?
The AEO Score Calculator is a structural assessment — it grades 20 signals on your page (schema, entity clarity, answer-format content, citation consistency) and returns a score out of 100. It tells you how machine-readable your site is. The AI Visibility Audit adds a second instrument: the AI Brand Perception Audit, which puts direct queries to ChatGPT, Perplexity, and Gemini to measure what each platform actually says about your brand right now. Client Zero ran both — 91/100 structural, 36/30/41 perception — and the gap between them is exactly what the audit is designed to surface.
How long does it take to see results from AEO?
Most businesses see meaningful structural improvement within 8–12 weeks when all four AEO Traction Stack layers are implemented. Entity clarity and schema markup take 1–4 weeks to fix. Answer-format content takes 3–6 weeks to produce. Citation consistency builds over 3–6 months and compounds over time. AI perception scores — what ChatGPT and Perplexity actually say about your brand — lag behind structural scores, because models refresh their knowledge on training cycles rather than in real time. That lag is normal. Starting the structural work earlier shortens it.
My AI visibility score is low — what should I do first?
The AEO Score Calculator shows your single highest-impact fix. For most low-scoring sites it is one of: missing or thin schema markup (add JSON-LD structured data), no FAQ content (add a FAQ section answering 5–10 customer questions), or inconsistent entity signals (check your business name and address match across Google Maps, Facebook, and directories). These three actions address the majority of low scores. AISearch Global's structural score started at 19/100 and reached 91 in six weeks by working through exactly this sequence. For a full prioritised action plan, the AI Visibility Audit delivers a 90-day roadmap.
Why can my website score highly on AEO but AI still not recommend my business?
Because structural score and AI perception are two separate measurements. A high structural score — like our own 91/100 — means the page is machine-readable and correctly formatted for AI to extract information from. It does not mean AI engines have ingested, weighted, or started citing that information yet. AI models have training cutoffs and citation refresh cycles that operate on their own schedules, independent of what you publish. The structural work makes citation possible. Building citation volume, external mentions, and content depth is what makes it happen. Both instruments — structural and perception — need to be measured to understand where you actually are.
How do I measure whether AEO is working?
The most direct measure is testing AI platforms manually: ask ChatGPT, Perplexity, Gemini, and Claude the questions your customers would ask, and note whether your business appears. Re-running the AEO Score Calculator every 90 days tracks structural score improvement over time. For deeper measurement, the AI Visibility Audit includes a baseline perception test — the same direct-query method that produced the 36/30/41 scores in this case study — which you can rerun after implementation to compare results across platforms.

Sources

Verify It Yourself — Free Tools, Independent Results

Every number on this page is reproducible using free, independent tools. We encourage you to run them — not to check our work, but to understand what your own site looks like from the outside. These are the same tools used in every AI Visibility Audit.

AEO Score Calculator

Enter any URL and get a live structural AEO score across 20 signals in under 60 seconds. No email required. For aisearch.global, the expected score is 91/100 — test it and see.

Run the AEO Score Calculator →

Google PageSpeed Insights

Test any URL at pagespeed.web.dev for Core Web Vitals, LCP, CLS, and performance scores on desktop and mobile. For aisearch.global: expected 98 desktop, 92 mobile.

Run PageSpeed Insights →

HubSpot Website Grader

Test any URL at website.grader.com for performance, SEO, mobile, and security scores. A useful cross-reference against the structural AEO signals — different methodology, same diagnostic intent.

Run HubSpot Website Grader →

Ask the AI Platforms Directly

Open ChatGPT, Perplexity, or Gemini and ask: "What do you know about AISearch Global, an AEO consultancy in Australia?" Compare what each says. The variation between platforms is itself the audit.

Open ChatGPT →

Then run the same tools on your own URL. That's where it gets interesting.

See your own gap before you guess at it

Same two instruments, your numbers instead of ours.

Client Zero remains a live experiment. Every quarterly update will be published publicly — whether the numbers improve or not.

Book your AI Visibility Audit