Graham Hyndman · Visibility Australia · September 2026
Every business owner I talk to has the same story. They've paid for at least one of these before — SEO, Google ads, a website rebuild, a review tool, a directory listing. Some have paid for four. And they still don't know whether any of it's working, because none of the people they're paying talk to each other.
That's not a marketing problem. That's a structural problem. And it's the reason I built Visibility Australia.
Your online visibility isn't one thing. It's six things, all changing at different speeds, all watched (or not watched) by different providers, none of them connected. When one breaks, the others compensate for a while — until they don't. And by the time you notice, you've been invisible for weeks and your competitor has been getting your calls.
Here are the six layers, what each one actually does, and why watching them together is the only approach that works.
Layer 1: Google Search
This is the one everyone thinks about. Where do you rank for the keywords that matter? Are you in the maps pack? Does your website show up when someone searches for what you do in your area?
Google Search is still the foundation. It's where the majority of business discovery starts, even in 2026. But it's no longer the whole picture — and it hasn't been for a while.
The problem with watching Google rankings in isolation: they shift constantly. Algorithm updates, competitor activity, seasonal changes, and now AI Overviews all affect where you appear. A monthly ranking report tells you where you were, not where you are. And if your SEO provider is only watching this layer, five other things are changing without anyone noticing.
Layer 2: AI Recommendations
This is the layer nobody was watching until recently — and it's the one that's changing fastest.
ChatGPT, Google Gemini, Perplexity, Claude, and Google's AI Overviews are now recommending businesses by name, with phone numbers, in response to natural-language queries. Not ranking them. Recommending them. There's a difference.
A ranking says "here are ten options." A recommendation says "call this one." The customer who gets a recommendation often never sees your website, never compares you to alternatives, and never knows you existed. They just call the business the AI named.
Research from Princeton and IIT Delhi, published at KDD 2024, identified specific content characteristics that influence whether AI platforms cite a source. Adding quotations from credible sources lifted visibility by 41%. Adding statistics lifted it by 31%. Keyword stuffing measured negative. The things that work for AI recommendations are different from what works for traditional SEO — and most providers haven't updated their methodology.
A separate study from Discovered Labs, analysing 2 million citations across four AI engines, found that prompt-content alignment — whether your page content matches the way customers actually phrase their questions — is the single strongest page-level factor, three times more influential than the next strongest.
If you're not monitoring what AI platforms say about your business daily, you're missing the fastest-moving layer of your visibility.
Layer 3: Google Business Profile
Your GBP is probably wrong right now. Not dramatically wrong — subtly wrong. Wrong enough that it's costing you without you knowing.
The Whitespark 2026 Local Search Ranking Factors survey, drawing on the expertise of 47 local SEO practitioners, ranks your primary GBP category as the number one local pack factor — ahead of physical proximity to the searcher. That means having the wrong category selected is more damaging than being further away from the customer.
Five of the top negative local ranking factors are GBP-related: incorrect categories, inconsistent NAP data, missing business hours, keyword-stuffed business names, and misconfigured service areas. All are fixable. Most businesses don't know they're there.
GBP also feeds directly into AI recommendations. When Google's AI Overview generates a local business recommendation, it pulls from your GBP data — categories, reviews, photos, service area. Stale or inaccurate GBP data doesn't just hurt your maps ranking. It makes the AI less likely to recommend you.
Layer 4: Reviews
Reviews are the layer that compounds when they're working and collapses when they're not.
Most business owners think about reviews as a star rating. That matters, but it's not the whole story. AI platforms and search algorithms also look at review velocity — how recently reviews have been posted, how frequently they arrive, and whether the pattern looks organic.
A business with 200 reviews but nothing in the last three months looks dormant to both Google and AI platforms. A business with 40 reviews but five new ones this month looks active and growing. The pattern matters as much as the total.
Review orchestration — systematically asking every customer for a review, without gating or incentivisation — is the only approach that's both effective and legal under Australian Consumer Law. The ACCC has been clear: selectively asking only happy customers, offering rewards for reviews, or discouraging negative reviews are all prohibited. The compliant approach is to ask everyone, equally, every time, and let the reviews reflect reality.
If your reviews have gone quiet, that's a visibility problem that affects every other layer. Google uses review signals in local rankings. AI platforms use them in recommendations. Directories surface businesses with stronger review profiles. One layer affects all six.
Layer 5: Citations and Directories
This is the layer everyone sets up once and forgets about. Your business is listed on Yellow Pages, TrueLocal, maybe Yelp, maybe a few industry directories. The listing was created three years ago with an old phone number and your previous business address.
Citation consistency — having the exact same business name, address, and phone number across every directory — is now a top-five factor for AI visibility specifically. AI platforms cross-reference your entity data across multiple sources. When they find inconsistencies, their confidence in your business drops. Lower confidence means fewer recommendations.
The fix is mechanical: audit every listing, correct every inconsistency, fill every gap. It's not exciting work. But it compounds. Every consistent citation reinforces your entity data. Every inconsistency erodes it. And the erosion is silent — you'll never get a notification saying "TrueLocal has your old number and it's hurting your AI recommendations." It just happens.
Australian businesses typically need consistent citations across 50-plus directories to be competitive. Most have fewer than 10, and at least half of those have errors.
Layer 6: Attribution and Measurement
This is the layer that tells you whether any of the other five are working. And for most businesses, it's broken.
Research from Search Engine Land, tracking 51,000 AI Overview events, found that 22.4% of AI-driven traffic gets misattributed to "Direct" in Google Analytics. The traffic comes from an AI recommendation, but your analytics tool records it as a direct visit — as if the customer typed your URL into their browser.
This means most businesses are underreporting their AI-driven traffic by roughly a quarter. They're getting more value from their visibility than they can see. And because they can't see it, they can't optimise for it.
Fixing attribution requires detecting the fragment identifiers that Google's AI Overviews append to URLs and correctly categorising that traffic in your analytics. It's a technical fix, not a content fix — but without it, you're making decisions about your marketing based on incomplete data.
Why nobody watches all six
The reason is structural, not negligence. Each layer has traditionally been watched by a different provider:
Your SEO agency watches Google rankings. Your review platform watches reviews. Your GBP is managed by whoever set it up (often you, once, years ago). Your directory listings are managed by nobody. AI recommendations are monitored by nobody. Attribution is misconfigured by default and fixed by nobody.
No single provider has had a reason to watch all six — until now. AI has changed that, because AI platforms pull from all six layers simultaneously when deciding whether to recommend you. A weakness in any one layer reduces the AI's confidence across the board.
That's why daily monitoring across all six layers, with a single system tracking how they compound and interact, is the approach that works. Not because any individual layer is complicated. But because the connections between them are where visibility is won or lost — and nobody sees those connections without watching everything at once.
What this looks like in practice
Imagine you're a mortgage broker in Brisbane. Your Google rankings are solid — page one for your main keywords. But your GBP has the wrong primary category (you're listed as "Financial Planner" instead of "Mortgage Broker"). Your reviews have been quiet for two months. Your TrueLocal listing has your old mobile number. And nobody has ever checked what ChatGPT says when someone asks "best mortgage broker in Northside Brisbane."
On paper, you're doing well. One layer is strong. But the other five are either wrong, dormant, or unmeasured. And the AI platform assembling a recommendation from all six can't build a confident enough picture to name you.
Your competitor — whose website isn't as good as yours, whose Google ranking is lower than yours — gets named instead. Because their GBP is accurate, their reviews are fresh, their citations are consistent, and their content matches the way customers actually ask the question.
That's the gap. It's not dramatic. It's not obvious. And it's costing you money every day it stays open.
What to do next
Start by checking where you actually stand. The two-minute test: ask ChatGPT, Gemini, Perplexity, and Claude the question a customer would ask before calling a business like yours. See who gets named. See if it's you.
If it's not — and for most Australian businesses it won't be — you now know which gap to close first. Check your GBP for accuracy. Check your reviews for recency. Check your directory listings for consistency. Check your website content for prompt-content alignment.
Or get the full picture at once. The Forensic Audit runs twenty real prompts across four AI platforms, screenshots the results, grades you against all six layers, and gives you the prioritised fix-it roadmap. 48 hours. Everything evidence-based. Everything you can verify yourself.
Because being invisible online isn't a mystery anymore. It's a measurement. And now you can see yours.
Free scan, about a minute. Which of the six layers are working, and which are letting you down.
Run my free scan →