SEO & AI Visibility

Google gave you a list. AI gives you a verdict.

For twenty years, search worked one way. Someone typed a question, Google handed them ten blue links, they clicked around, compared options, and made up their own mind. Your job as a marketer was to win one of those clicks.

AI search skips all of that. Someone asks a question, and the AI reads the options for them, evaluates the sources, weighs them against each other, and hands back a conclusion. One answer. Maybe two or three businesses mentioned. Everyone else gets left out entirely, and the person asking never even knows you were an option.

If you think that traffic is too small to matter, look at the conversion data: visitors who arrive from AI recommendations convert about eight times more than traditional traffic, and they convert faster (NP Digital measurement, 2026). By the time they land on your website, the AI has already vetted you, compared you, and recommended you. They arrive pre-sold. This isn\'t a new traffic source, it\'s the highest-intent traffic that exists, and your rankings dashboard can\'t see any of it.

The question every part of this page hangs on: how do these systems decide who makes the cut?

Trust just became visible

For the first time, one of these systems is showing you its cards. Perplexity now labels the sources inside its answers across four public tiers: no label, credible, trusted, and authoritative. It even tells you why a source earned it: "primary source", "specialised industry publication", "strong track record in this field."

For twenty years, Google\'s algorithm was a black box. You never saw why a page ranked third instead of first. Now an AI engine is showing you its trust map on screen. That looks like a small product feature. It\'s more like a confession. These systems aren\'t just reading what a page says, they\'re running a background check on who said it. ChatGPT and Gemini are making the same judgment; they just don\'t put a badge on it yet.

What actually gets you recommended

NP Digital analysed 82 factors to determine what gets a brand recommended on ChatGPT. The order matters:

1. Relevancy, the obvious one.
2. Brand mentions, how often your name appears across the web on sites you don\'t own. Nearly tied with relevancy.
3. Reviews. Google, Trustpilot, Yelp, industry-specific platforms.
4. Authority, is anyone qualified attached to your name?

The most valuable currency of the old internet barely registers on the new one. Backlinks scored 1.9 out of 5. Nearly the bottom of the list.

What replaces it is evidence: real names on content, independent press, reviews, third-party citations, other trusted sources vouching for you. Expert authorship scored 4+ across every AI platform. If your smartest people have no public evidence attached to their names, the machine can\'t see your expertise. To the system, it doesn\'t exist. (Why AI trusts what others say about you, not what you say about yourself.)

AI is writing your first impression. Right now. Without you.

Think about everything that exists about your business online. Your website. Your founder\'s LinkedIn. Old interviews. Podcasts. Customer reviews. Press from five years ago. Reddit threads you\'ve never read. AI pulls all of it together in seconds and compresses it into one paragraph. And that paragraph is becoming the first thing a potential customer learns about you, before they visit your site, before they click an ad, before they talk to sales.

Twenty years of building a business, summarised in four sentences you\'ve never read and can\'t edit. That\'s the part that should keep you up at night.

Compression is brutal. Whatever survives the compression becomes the story. For most businesses, what survives is a mess. The website says one thing, the founder\'s profile says another, the press coverage is about a product you pivoted away from three years ago, the team has deep expertise but there\'s zero public evidence connecting their names to it. A human might untangle that. The machine doesn\'t. Muddy signals in, muddy paragraphs out, or worse, a confident paragraph that\'s flat wrong about what you do.

Your ads read from the same source material now

If you think this only matters for AI search, look at what\'s happening inside the platforms you\'re already paying. Google\'s AI-powered ad products don\'t just distribute your ad any more, they read your website, your landing pages, your product information, your messaging. Then the system uses all of that to decide what your business offers, who to match it with, and what your ad should even say.

Your website used to be the destination an ad pointed to. Now it\'s the source material the machine studies before it builds the campaign. That means every inconsistency between your channels, website, social, PR, reviews, is noise in the machine\'s understanding of what business you\'re in. A confused machine matches you with the wrong audience, generates the wrong message, and recommends your competitor to your customer.

What we do: build the reputation the machines can read

Byron Sharp\'s work on mental availability, being the brand that comes to mind when a buyer enters the category, maps directly onto this. In the AI era, "coming to mind" increasingly means "being the brand the AI recommends." Jenni Romaniuk\'s Category Entry Points tell us which moments to target. A plumber isn\'t just competing for "plumber near me." They\'re competing for the moments that trigger the thought: "the tap is leaking", "we\'re renovating the bathroom", "the hot water system died." Each of those is a query someone types into ChatGPT. Each is a content opportunity. (Why most buyers don\'t think about your business.)

The strategy isn\'t "publish more." Publishing more of a contradictory story just makes the contradiction louder. The play is fewer, stronger signals, every piece of proof reinforcing the same story, easier for a machine to verify from any angle. Six components:

1. A website that says exactly who you are

Unambiguous positioning. Clear service definitions. Server-side rendered so AI crawlers can actually read it (Aleyda Solis\'s research: AI crawlers skip client-side JavaScript). Complete structured data. Organization, LocalBusiness, Person, Article, FAQ schema. Consistent NAP across every listing.

2. Content that establishes what you believe

Not commodity summaries. Google\'s own AI Optimization Guide (May 2026): "Content that simply restates what every other site already says is unlikely to be selected." The same applies to LLMs. We build point-of-view content that AI systems cite because you\'re the only place they can get it from. (Why generic content gets ignored by AI.)

3. Real names attached to real expertise

Bylines on every article. Author schema linking your name to your LinkedIn. Speaking, podcast appearances, quoted expertise. Expert authorship scored 4+ across every AI platform NP tested, no bylines = invisible expertise.

4. Case studies that document what you\'ve done

Real client work, real numbers, real names (with permission). AI systems cite specifics. "We grew Client X\'s Google Ads revenue 40% in six months" beats "we deliver results."

5. Reviews that prove the experience

Google Business Profile reviews first (heavy AI weighting for local queries). Then Trustpilot, industry-specific platforms, video testimonials. Sites with active review profiles have ~3× higher AI citation probability in category queries.

6. Original research that gives others a reason to cite you

Cyrus Shepard\'s analysis of 400+ sites: 92.9% of sites winning Google traffic own proprietary data assets (original research, unique databases, community content). AI systems cite what only you can source. This is the highest-leverage, most-neglected component for SMEs.

Old SEO vs the reputation approach

How this connects to everything else we do

Mark Ritson argues for bothism in marketing: brand and performance are complements, not opposites. Reputation is the same. It\'s the layer underneath everything:

• Google Ads, your reputation now feeds the AI that builds the campaigns. A cleaner reputation = better ad matching and lower CPA.
• Meta Ads, creates demand. When people see you on Instagram then ask ChatGPT about your category, both signals feed the same verdict.
• Email Marketing, captures the visitors reputation-work brings in. A blog reader who joins your list becomes a lead you can nurture without paying for another click.

Binet and Field\'s 60/40 rule applies: reputation work is the long-term brand-building 60%. It compounds. The businesses that invest consistently are the ones that eventually don\'t need to spend as much on ads.

Your homework

Before you do anything else, before you book a call, before you hire anyone, ask ChatGPT to describe your business. Then Perplexity. Then Gemini. Read the paragraphs carefully.

If you don\'t like the verdict, or you don\'t want to figure this out alone, that\'s what we do. Get in touch, we\'ll run the verdict test on your business, show you the specific signals each engine is picking up, and build the reputation system to fix it.

Thousands of your future customers will hear that verdict before you ever get in a word. Most of your competitors have never read theirs. The ones who fix their verdict first are going to be very hard to catch.

Dream Outcome is an Australian digital marketing agency helping SMEs grow through Google Ads, Meta Ads, SEO, and Email Marketing.

Frequently asked questions

What's the "verdict" and how do I read mine?

Ask ChatGPT (or Perplexity, or Gemini): "Tell me about [your business name] in Adelaide." Read the four-sentence paragraph it returns. That paragraph is what many of your future customers now hear before they visit your website. Most business owners have never read theirs.

How long does SEO / AI visibility take to work?

For a site with reasonable technical foundations, expect meaningful movement in 3-4 months and material commercial traffic in 6-9 months. AI citations often move faster than Google rankings, sometimes within weeks, because AI systems weight brand mentions and third-party citations more than traditional SEO signals.

Isn't this just "SEO for AI"? Do I need three separate strategies (SEO / AEO / GEO / LLMO)?

No. The market wants to sell you three strategies. The reality is one: build a coherent, verifiable reputation across every public surface. The same fundamentals (clear positioning, expert authorship, third-party citations, structured content) win on Google, ChatGPT, Perplexity, and Gemini. We call it SEO because that's what buyers search for. The method is unified.

Do backlinks still matter?

Less than they used to and less than most SEOs will tell you. NP Digital's 82-factor ChatGPT ranking study scored backlinks 1.9 out of 5, nearly the bottom of the list. What replaced them: brand mentions, reviews, third-party citations, and expert authorship. If someone's selling you a link-building package as the answer, they're selling you the 2020 playbook.

Do I still need SEO if I run Google Ads?

Yes, and it now matters for a second reason. Google's AI-powered ad products read your website, landing pages, and messaging as source material to decide what your business offers, who to match it with, and what your ads should say. Your reputation now shapes both organic visibility and paid campaign performance. They can't be run in isolation any more.

How do you know if it's working?

We measure four things: (1) the verdict, how AI systems describe your business, tracked monthly; (2) brand mentions across independent sources; (3) organic traffic to money pages plus lead/revenue attribution; (4) commercial-intent keyword rankings as a diagnostic. Rankings alone aren't the goal, the goal is the verdict.