How to Rank #1 in 2026: Three Steps That Actually Work for SEO and AEO

If you've watched your organic clicks slide this year even though your rankings haven't moved, you're not imagining it. SEO has changed more in the last twelve months than in the previous decade, and most of the playbooks people are still working from are quietly broken.

The good news is that the new model is actually clearer than the old one. There are three steps, the evidence behind them is solid, and you can start applying any one of them this week.

This piece is the version I wish I'd had on my desk in January.

What changed in 2026

Two numbers do most of the work in setting up why the playbook has shifted.

The first is the click-loss data. When an AI Overview appears in search results, the #1 ranking page now loses 58% of its clicks, according to Ahrefs' April 2026 refresh of their original study. Twelve months ago, the same number was 35%. So the impact has nearly doubled in a year, and AI Overviews are still expanding into more query types.

The second is the correlation data. Ahrefs ran a study of 75,000 brands in May 2025 to work out which signals actually predict whether a brand appears in AI Overviews. The result is the cleanest evidence anyone has produced so far that the old ranking factors no longer translate cleanly to the new surface:

Two things to notice. First, branded mentions — not links — sit at the top. Second, the gap between #1 and the link-based factors is enormous: roughly three times the predictive power of raw backlink count. The 75-to-100% top-quartile brands get 169 AI Overview mentions on average; the next quartile down gets 14. There's a 10x leverage curve, and 26% of brands have zero AI Overview mentions at all.

This isn't vendor speculation. Rank Math cited the Ahrefs study by name in their own 2026 playbook, and Google's 2024 Content Warehouse API leak — confirmed by sworn DOJ testimony — separately validated that user-behaviour signals (the leaked NavBoost system, including goodClicks, badClicks, and lastLongestClick) are primary ranking drivers. There is a real consensus forming, even if it hasn't trickled down to most agency proposals yet.

The other thing the consensus has surfaced is the action-query carve-out. AI Overviews are good at summarising information. They can't take action. So queries like "backlink checker", "perth tax accountant", "power automate consultant", "BAS calculator" — anything transactional, local, or tool-based — almost never trigger an AI Overview. Traditional SEO wins those queries the same way it always did.

That gives us the shape of the new playbook. Three steps, each tackling a different surface.

Step 1 — Match search intent, then upgrade what you already have

The most common mistake we see in 2026 is the same one most agencies have been making for a decade: defaulting to "write more content" the moment traffic dips. For most established websites, that's the expensive option. The cheap option is sitting in your Google Search Console right now.

The intent check before anything else

Before you write a word, look at the actual SERP for your target keyword. Then ask two questions.

The first is format. What does Google currently rank? If the top results are interactive calculators or product category pages and you're planning a 2,000-word blog post, you've misread the intent and you're going to lose. If the top results are long-form guides and you're planning a thin product page, same problem. Match the format Google has chosen before you decide what to build.

The second is the AI Overview filter. Run the query. Does an AIO appear? If yes, expect the page to lose more than half its clicks even if it ranks #1. That doesn't mean you don't target it — it means you have to be honest with yourself about what success looks like. If the query has high commercial value and an AIO appears, you're now competing for AI citation as much as for the link.

The GSC audit that beats publishing new content

Once you've sorted intent, the highest-ROI work isn't a new article. It's a Google Search Console audit, and it takes about an hour.

Open Google Search Console, go to Performance, set the date range to the last 28 days, and make sure average position is showing in the table. Then sort your pages by average position and look for two specific patterns.

Pages ranking positions 4 to 15 with high impressions. These are pages Google has already decided are almost good enough for page one. They need a push, not a replacement. Update the content, refresh the publish date legitimately, tighten the title tag, add the questions people are actually searching, and improve the internal linking. Most of these pages will move up 3 to 8 positions within 4 to 8 weeks.

Pages that have dropped from 4-15 into 16+. Add the previous-period comparison in GSC. If a page used to rank #8 and now ranks #22, recover it. Usually the fix is competitive — someone else updated their version more recently. Refresh the content, add what the new top result has that yours doesn't, and resubmit to GSC for re-indexing.

The third pattern, less obvious but just as important: zero-traffic dead pages. Pages that have been indexed for over a year, attract no clicks and no impressions, and have no internal links pointing to them. These pages don't just sit there harmlessly. Google evaluates your site holistically — the leaked Content Warehouse documentation made this explicit — and a domain full of dead weight drags down the pages that are working. Either rewrite them to genuinely serve a query, consolidate them into a stronger page with a 301 redirect, or remove them.

We've run this audit on every client retainer in the last six weeks. Without exception, the existing-content path has produced more measurable traffic gain than the new-content path in the same window.

Step 2 — Build deep topical authority with hub-and-spoke

The second shift in 2026 is about how Google decides whether to trust you on a subject at all.

The pre-2024 model rewarded broad sites that wrote about anything. The post-2024 model — and this is where the leaked algorithm documentation, Rank Math's playbook, and our own client data all agree — rewards sites that go deep on a narrow set of topics and ignore everything else.

The structural pattern is called hub-and-spoke, and it's the single biggest content lever still available to small and medium businesses in 2026.

How the architecture works

You start with a topic you want to own. Not "marketing" — too broad. Not "email marketing" — still too broad. Something like "email automation for Australian e-commerce stores" or "Power Automate flows for trades businesses". Narrow enough that a focused 30-article body of work can plausibly cover every angle.

Then you write 20 to 30 spoke articles, each covering one very specific sub-question. "How long should a welcome email sequence be?" is a spoke. "What's the best email automation tool for a Shopify store under 1,000 customers?" is a spoke. "How to handle GST on automated receipts in Australia" is a spoke. Narrow, specific, answerable in 800 to 1,500 words each.

Every spoke contains a contextual internal link pointing back to the pillar hub — the broad page that anchors the whole cluster.

Counterintuitively, you write the spokes first. The pillar is too broad to rank on its own at the start. The spokes pick off long-tail queries individually, and as they start to rank, they pass authority up to the pillar via the internal links. After three to six months, the pillar itself starts ranking on the broad head term. By that point Google's crawlers have read your site as a tightly-woven web of relevance, and you've earned what the industry calls topical authority — the algorithmic equivalent of being the recognised expert on the subject.

If your site covers travel tips on Monday, personal finance on Wednesday, and software reviews on Friday, this won't work. Google will not trust you to rank for any of them. Pick a topic and own it.

Prove first-hand experience

The other thing the 2024 Google leak made explicit is that first-hand experience signals matter more than they used to. Google's quality rater guidelines added the second "E" to E-A-T — Experience — for a reason. AI-generated content can fake expertise. It can't fake having done the thing.

So when you write, prove you've done the thing. Use your own screenshots, not stock photos or vendor press kits. Show the intermediate steps, not just the final result. Mention the specific error messages you hit, the workarounds you found, the sensory details only someone who did the work would notice. These are the nuances that purely AI-generated content cannot fake, and they're disproportionately what AI models pull from when they choose source material.

Cite outbound sources

The other myth worth killing is the idea that linking out to other sites "leaks PageRank". It doesn't, and it never has. Linking out to authoritative relevant sources — Australian Bureau of Statistics data, official Microsoft documentation, AS/NZS standards, Google's own developer guides — actively validates your content. AI models pull from pages that cite credible sources at noticeably higher rates than from pages that don't.

We rewrote two service pages in late April adding outbound citations to AU government data and Microsoft Learn modules. Both pages picked up impressions in the following three weeks. The link equity loss was, predictably, zero.

Step 3 — Earn brand mentions and external trust signals

The third step is where the new model diverges hardest from the old one, and it's the step most businesses haven't started.

The Ahrefs study makes the case bluntly: branded web mentions correlate with AI Overview visibility at 0.664. Backlinks correlate at 0.218. If you want to be cited by AI engines — Google AI Overviews, ChatGPT Search, Perplexity, Gemini — you need your brand name appearing on the pages those engines pull from when they fan out a query.

There are four substrates that produce these mentions, and most businesses only need to pick two.

Original-data linkable assets

If you have proprietary data — sales numbers, customer surveys, telemetry, repair statistics, anything quantitative that you collect as a normal course of business — you have the raw material for the most effective link and mention earner in 2026. Turn it into a short report. Run a survey of your own customer list. Analyse public open-source data in a way nobody else has. Build an embeddable chart or interactive tool.

Then pitch it to local news outlets, niche bloggers, and trade publications. The story isn't "we exist". The story is "here's a data point about your industry that you can use in your own piece". Real estate firms have done this for years with house-price reports. The technique works for any sector that touches numbers.

Digital PR and source platforms

Australian-friendly journalist-source platforms like SourceBottle and Qwoted give you a direct line to journalists looking for expert sources. The trick is not pitching your product. The trick is answering the question with concise, useful expertise — what economists call "high-quality public goods" — and earning a quote with a link back. Done consistently, it builds your reputation as the named expert in your space, and your name starts appearing in articles AI engines pull from when they fan out queries on your topic.

This is one of the highest-leverage mention engines available to a business without a six-figure PR budget.

Community participation

Niche subreddits, specialised industry forums, LinkedIn groups with engaged readers, podcast appearances on shows your prospects actually listen to. These are slow substrates — you can't fake authentic participation in a community for very long — but they're the ones AI models index most aggressively. Google AI Overviews in particular pulls from Reddit and YouTube at rates wildly disproportionate to those platforms' share of overall web traffic.

If you sell to a niche, find the three places that niche gathers online, and become a recognisable contributor.

Direct outreach to listicles, comparisons, and review pages

The final substrate is the most tactical. AI engines fan out queries like "best [product] in [city]" and "[brand] vs [brand]" by pulling from listicles, comparison articles, and review sites. If your business is mentioned on those pages, you appear in the fanned-out answer. If you're not, you don't.

The work here is straightforward but unglamorous. Find the listicles and comparison pages that already rank for your category. Reach out to the authors with a useful angle — a free sample, an interview, a data point they can use. Build the placements one at a time. This is the tactic Ahrefs prescribes most directly in their own playbook, and it works because the asymmetry favours small businesses: the journalist or blogger only needs one good reason to add you to their list.

The honest version

These three steps tile cleanly onto each other. Step 1 protects what you already have and produces measurable gains in 4 to 8 weeks. Step 2 builds the topical authority that wins on traditional rankings and stops the slow bleed against bigger competitors. Step 3 is the long game — 6 to 12 months to compound — but it's the only one that unlocks AI Overview visibility at the scale that matters.

If you're an Australian small or medium business with no SEO team, the realistic starting point is Step 1. Run the GSC audit this week. Pick three pages ranking 4-15 with high impressions and upgrade them. You'll see movement before you've finished planning Step 2.

We do this work for a portfolio of SEO clients across Australia, and the framework above is exactly what we run on those retainers. The principles work whether you're publishing through our hand, your own team, or a different agency. The cost of ignoring them grows every quarter the AI Overview share keeps expanding.

If you'd like a look at how this would apply specifically to your site — which pages to upgrade first, which topical authority gaps to close, where the brand-mention substrate would compound fastest — book a free consultation and we'll map it out.

Either way, start with Step 1. The biggest mistake in 2026 SEO isn't picking the wrong tactic. It's still publishing more content when the cheaper win is sitting in your existing pages.