Your Website Isn't a Source. That's Why Google and AI Ignore It.
Here's a question most business owners have never considered: if ChatGPT, Google's AI Overview, or Perplexity were writing an answer about your industry right now, would they have any reason to cite your website?
Not link to it. Not rank it. Cite it. Quote it as a source the way a journalist would quote an expert.
For the vast majority of Australian SMEs, the honest answer is no. And that single fact explains more about your search visibility than any technical SEO audit ever will.
The rules of search have shifted beneath our feet. Rand Fishkin's latest SparkToro data shows that 68% of all Google searches in 2026 now end without a click. That's up from 60% just two years ago. Meanwhile, an Ahrefs study of 863,000 keywords found that only 38% of pages cited in Google AI Overviews also rank in the traditional top 10. The other 62% come from elsewhere, including pages that don't rank in the top 100 at all.
The game isn't "rank higher" anymore. It's "be worth quoting."
The Data That Should Make Every Small Business Pause
Cyrus Shepard, founder of Zyppy SEO, spent 12 months analysing more than 400 websites to identify what separates the sites gaining Google traffic from those losing it. He measured five features using Spearman correlation coefficients, and the results paint a stark picture.
| Feature | Correlation with Traffic Growth | What It Means |
|---|---|---|
| Product or Service | 0.391 | You sell something real, not just information |
| Task Completion | 0.381 | Visitors can actually DO something on your site |
| Proprietary Assets | 0.357 | You own data, tools, or content nobody else has |
| Topical Focus | 0.250 | You go deep on your niche, not wide across everything |
| Brand Demand | 0.206 | People search for you by name |
The win rates tell the real story. Sites with four or five of these features achieved win rates of 68-70%. Sites with none? 13.5%.
Look at the third row. Proprietary assets. Not backlinks. Not keyword density. Not publishing frequency. Owning something that other sites can't easily replicate: unique databases, original research, calculators, proprietary benchmarks, first-hand case studies with real numbers.
This is the feature that separates a website that gets cited from one that gets ignored. And it's the feature almost no SME has thought about building.
Why Most SME Websites Are Invisible to AI
Think about what's on your website right now. If you're like most small businesses, it's some version of:
- A homepage explaining what you do
- A services page listing your offerings
- An about page with your story
- A blog with articles that rehash information available on a hundred other sites
AI systems don't cite opinions. They cite sources.
The AI Platform Citation Source Index 2026, which analysed over 680 million individual citations across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews, found that the top 15 domains capture roughly 68% of all AI citation share. Reddit alone appears in approximately 40% of generative answers. Why? Because Reddit contains something most business websites don't: real people sharing real experiences with specific details.When someone asks an AI "What's a good plumber in Adelaide?", the AI isn't going to cite your website that says "We're Adelaide's trusted plumbing experts." It's going to cite the Reddit thread where a real person wrote "Used [business name] last week, they quoted $380 for a blocked drain and showed up within 2 hours."
The specificity is the value. The first-hand data is the value. The thing that can't be copied from a competitor's website is the value.
Rory Sutherland's Costly Signal: Why Investing in Original Content IS the Marketing
This is where behavioural economics explains what the data shows.
Rory Sutherland, Vice Chairman of Ogilvy UK, has spent decades arguing that the meaning and significance we attach to something is often proportional to the expense with which it's communicated. He calls this costly signaling, borrowing from evolutionary biology. A peacock's tail is expensive to grow and maintain. That's the point. The cost IS the signal. It proves genetic fitness precisely because it can't be faked cheaply.
Sutherland's most famous example: a FedEx letter is assumed to be important because "who would send something unimportant for eight dollars?" The medium signals the message's value before you even open it.
Now apply this to your website content.
When a plumbing business publishes "5 Tips to Prevent Blocked Drains" and the article contains the same generic advice found on 200 other plumbing websites, what signal does that send? That you spent five minutes (or asked ChatGPT for 30 seconds) producing something with zero original value. The cheapness of the signal communicates the cheapness of the investment.
But when that same plumbing business publishes "We Analysed 847 Adelaide Callouts From 2025. Here's What Actually Blocks Your Drains" with a breakdown of the most common causes by suburb, season, and pipe age, backed by their own job data, that's a costly signal. It communicates:
- We have real experience (847 callouts means we're busy)
- We track our work (suggesting professionalism and systems)
- We're willing to share what we know (building Cialdini's reciprocity)
- We can't be easily copied (your competitors don't have your data)
As Sutherland puts it: "It is much easier to be fired for being illogical than it is for being unimaginative. The fatal issue is that logic always gets you to exactly the same place as your competitors." Every business following the same SEO playbook (publish 4 blog posts a month, target long-tail keywords, add FAQ schema) lands in exactly the same place. The businesses that invest in creating something genuinely original end up somewhere nobody else can reach.
What "Proprietary Assets" Actually Look Like for a Small Business
This is where most business owners check out. "I'm not a research institution. I don't have data scientists. I'm a tradie / accountant / physio / manufacturer."
You don't need a research department. You need to recognise that you already own data nobody else has. You just haven't packaged it.
| Business Type | Data You Already Have | Proprietary Asset You Could Create |
|---|---|---|
| Trades (plumber, electrician, HVAC) | Job logs, quotes, seasonal demand patterns | "Average repair costs in [city] 2026" pricing guide with real numbers |
| Professional services (accountant, lawyer) | Client questions, common mistakes, outcome data | "The 10 mistakes we see most in [specific niche] tax returns" with frequency data |
| Health & wellness (physio, dentist, PT) | Patient outcomes, treatment durations, recovery data | "Average recovery timeline for [condition]: what our 200+ patients experienced" |
| Home services (cleaning, landscaping, pest control) | Service frequency data, seasonal patterns, pricing | Interactive quote calculator based on real job data |
| B2B services (marketing, IT, consulting) | Campaign results, project outcomes, industry benchmarks | Annual industry benchmark report for your specific niche |
Notice the pattern. None of these require new systems or expensive tools. They require looking at the data you already collect through normal business operations and packaging it into something genuinely useful. Lily Ray, VP of SEO Strategy at Amsive, has been arguing that E-E-A-T principles have evolved into a universal trust framework for the entire AI ecosystem. Experience, Expertise, Authoritativeness, and Trustworthiness aren't just Google's quality guidelines anymore. They're the criteria every AI system uses to decide whether your content is worth surfacing to users. And nothing demonstrates Experience more convincingly than first-hand data from your actual work.
The Information Gain Principle: Why Saying the Same Thing Differently Isn't Enough
Google holds a patent for "information gain scoring" that describes a mechanism for evaluating how much new information a document adds beyond what the user has already seen. While we can't confirm exactly how Google implements this, the principle aligns perfectly with what Shepard's data shows: sites that add unique information to a topic win. Sites that repackage existing information lose.
This is why the standard content marketing playbook has stopped working for most SMEs. "Write a blog post about [topic your audience cares about]" produces content with near-zero information gain when a thousand other businesses have written the same blog post about the same topic.
The shift required is fundamental. It's not about writing better versions of existing content. It's about creating content that contains information that doesn't exist anywhere else on the internet.
We've written before about why AI makes most marketing faster but not better. This is a perfect example. AI can help you produce a polished 2,000-word article about "how to choose an accountant in Adelaide" in minutes. But if that article contains nothing your competitors' AI-generated articles don't also contain, you've produced content faster while adding zero value to the internet. Google knows. The AI models know. Your rankings reflect it.
The Citability Flywheel: How This Compounds Over Time
Here's what makes proprietary assets different from every other SEO tactic: they compound.
Ethan Smith, CEO of Graphite, has spent 18 years in SEO and now leads the most cited work on Answer Engine Optimisation. His research shows that LLM traffic converts 6x better than Google search traffic, because users who arrive through AI assistants have built up significant intent through conversation and follow-up questions. They're not browsing. They're buying.When your business becomes a cited source in AI answers, something interesting happens:
- AI cites your original data in an answer to a user's question
- The user visits your site because the AI linked to you as the source
- Other content creators reference your data in their articles, building backlinks
- Google sees increased brand searches for your business name (Shepard's "brand demand" factor)
- More AI systems discover and cite you, because you now have both original data AND third-party validation
Think about it through Byron Sharp's lens. Sharp's research across 130+ brands proves that brands grow by being easy to think of and easy to buy. He calls these mental availability and physical availability. In search, your content IS your physical availability. If your website contains nothing an AI would cite or a journalist would reference, you have no physical availability in the fastest-growing discovery channel.
And Shepard's data on brand demand confirms Sharp's framework directly. Brand demand doesn't come from publishing branded blog posts. It comes from consistent market presence through original research, distinctive products, memorable frameworks, expert commentary, useful tools, and content that people refer back to. Search traffic without brand growth is often rented attention. Search traffic that increases brand demand becomes an asset.
What This Actually Costs (And Why It's Cheaper Than You Think)
The irony of costly signaling is that it doesn't have to cost a fortune. It has to cost enough that it can't be easily faked. For most SMEs, the real cost is time and willingness, not money.
The cheap way that doesn't work: Pay a content writer $200 to produce four blog posts per month that say the same thing as everyone else's four blog posts. Annual cost: $2,400. Annual result: negligible. The "costly signal" way that compounds: Spend one afternoon per quarter pulling data from your existing business operations. Package it into one genuinely original piece. Annual cost: four afternoons of your time plus perhaps $500 in design. Annual result: four pieces of content that no competitor can replicate, that AI systems can cite, and that build your authority over time.We've seen this play out with clients. We wrote about the post-click gap and how marketing dashboards mislead using real campaign data that generic marketing blogs simply don't have access to. The articles that perform best, without exception, are the ones built on specific data that can't be found elsewhere.
Aleyda Solis, one of the most respected technical SEOs working today, puts it clearly: the winning approach is ensuring crawlability of key on-brand content that targets the full customer journey, establishes a clear USP, and aligns with the third-party platforms where AI citations actually originate. The technical foundation matters. But without original content worth crawling, the best technical SEO in the world is just making an empty building easier to find.The Three Questions That Reveal Whether Your Site Is a Source
Before you commission another batch of blog posts, ask yourself:
1. Does our website contain any data point that doesn't exist on a competitor's site? If the answer is no, you're producing content with zero information gain. Every page is competing against every other page saying the same thing. Start with one original data point from your business operations. 2. If an AI were writing a definitive answer about our industry in our city, would it have any reason to reference us? If not, you're invisible to the fastest-growing discovery channel. Your goal isn't to trick AI into citing you. It's to create something genuinely worth citing. 3. Could we publish something this quarter that a journalist, an industry association, or a university could reference? This is the "costly signal" test. If your content wouldn't pass editorial review at a trade publication, it won't pass the increasingly sophisticated editorial judgment of AI systems either.What This Means for Your Business
The SEO landscape of 2026 has split into two lanes.
Lane one: Businesses producing high volumes of generic content, fighting for scraps in a world where 68% of searches don't generate clicks and AI systems cite only the sources they trust. This lane is getting more crowded and less rewarding every month. Lane two: Businesses that have stopped trying to rank for keywords and started trying to become sources. They publish less but own more. Every piece contains something original. Their content compounds because it's cited, referenced, and linked to, not because it was optimised for a keyword.The data from Shepard, Fishkin, Ray, and Smith all point to the same conclusion: the businesses winning in search are the ones that have something worth finding.
You don't need more content. You need content that matters.
Further Reading
- 5 Data-Backed Features of Websites Winning Google in 2026 - Cyrus Shepard's analysis of 400+ sites, the most rigorous study of what drives Google traffic growth right now
- In 2026, Less Than One Third of Google Searches Still Send a Click - Rand Fishkin's latest zero-click data showing 68% of searches end without a click
- Only 38% of AI Overview Citations Pull From the Top 10 - Ahrefs study of 863,000 keywords showing AI Overviews cite far beyond traditional rankings
- The Ultimate Guide to AEO - Ethan Smith on Lenny's Podcast explaining how to get cited by AI answer engines
- The AI Search Optimization Checklist - Aleyda Solis's practical framework for AI search readiness
Dream Outcome is an Australian digital marketing agency helping SMEs grow through Google Ads, Facebook Ads, and Email Marketing.