How an SEO agency should deliver: The 5 layer method

A practical look at how an SEO agency delivers SEO, from strategy and technical foundations through to content, authority and ongoing optimisation.

SEO briefs typically start with symptoms

Nobody hires an SEO agency because they want a keyword spreadsheet or a technical audit report. The brief always arrives as a symptom, not a diagnosis and it usually sounds like one of these:

• We used to rank for this and now we don't
• Our competitor is everywhere and we're nowhere
• We're getting traffic but not enquiries worth having
• We're not showing up when people ask ChatGPT for a recommendation
• We're launching into a category where nobody knows us yet

None of those has a single fix behind it and each one is what shows up on the surface when something underneath isn't working.

Fixing one thing rarely fixes all rankings

A site that's slipped is almost never fighting one problem while the search landscape itself has changed. It's no longer just the Google results page, it's also AI Overviews, chatbot answers, comparison sites and reviews. Getting found now depends on multiple things done well, not one.

The five things that make up good SEO

Each one builds on the one before it, which is why we've listed them in this order. In practice we don't tackle them one at a time. Technical work usually starts while research is still underway, reviews and reputation are handled continuously from day one and new content gets written for the pages already underperforming, as well as new pages where we spot a gap.

Layers one to four are the work. Layer five is what tells you whether it's working and which of the other four needs attention.

A note on wording: We say "enquiries" throughout as shorthand, since that's the goal for most service businesses, but if you're an eCommerce business, read that as sales.

Layer 1: Getting your website technically sound

Everything else matters less if Google and AI tools can’t properly read your site in the first place. It's not glamorous work, but it's one of the most common reasons a website underperforms.

What this involves

• An audit of your site to find anything stopping Google from reading it properly
• A site structure that matches how customers think about your business
• Fixing broken links and outdated pages, especially after a website redesign
• A faster, smoother experience on both desktop and mobile
• Making sure your site is accessible to everyone, including people who use screen readers
• Getting important information out of PDFs and onto proper web pages
• Making sure FAQs and other accordion or tab content are actually in the page's code when it loads, not just added in after a click
  • PDFs can rank, but important content is usually better placed on the page itself where search engines and AI tools can access it more easily.
  • Accordions and tabs are fine too, as long as the content is already present in the page code when it loads. If it only appears after someone clicks, search engines and AI tools may not see it.

Google may tolerate a messy site because your brand has history and authority. AI tools like ChatGPT are less forgiving: if they can’t quickly understand a page, they may favour clearer sources instead.

You're probably feeling this if:

• Your visibility dropped after a redesign or a website move
• Old pages that shouldn't exist any more still show up in search
• Your site has felt steadily slower over the past year or two
• Important information like pricing, policies or specs, only exists in PDFs rather than web pages
• Your FAQs only get added into the page after someone clicks to open them, rather than being in the code from the start

Layer 2: Understanding what customers search for

Most businesses base their content on the handful of terms they already know they want to rank for. That's a fine place to start, but a poor place to stop.

What this involves

• Working out exactly what your customers type into Google, and increasingly, what they ask ChatGPT and similar tools
• Understanding how you compare to competitors in the searches that matter
• Capturing the real questions and pain points customers raise, not the ones we assume
• Checking branded names still include the everyday terms customers search for, such as “Lush nail polish” rather than just “Lush”.
• Turning all of that into a clear plan for what to write next and in what order

The way people search is changing. ChatGPT queries are longer and more specific than Google searches, so relying on traditional keyword research can miss how customers are actually looking for you.

You're probably feeling this if:

• Your content hasn't materially changed in a couple of years
• Content decisions get made by opinion, or by whoever asked most recently
• You have no real sense of what customers ask before they know the right words for what you do
• Your product or service names rely on internal branding without the everyday term attached, so people searching for what you sell never find you

Layer 3: The content and pages themselves

This is where most of the budget goes and where it's most often wasted. The instinct is to publish more. The better move is usually to fix and fill in what's missing before writing anything new.

What this involves

• Rewriting key pages to lead with the problem you solve, not just your features
• Building the pages customers need to decide, including comparisons, use cases and FAQs
• Answering common sales questions so the website does some of the work
• Making your site easier to read and easier to act on

Freshness matters here, but not evenly across every page. Google leans towards newer or recently updated content for things that genuinely change, current pricing, "best of" lists, anything tied to a year or a fast-moving trend. 

For an evergreen page explaining what something is, the exact publish date barely matters as long as the information's still accurate. AI tools lean into freshness even harder than Google does, so a page left untouched for years is less likely to get picked up in an AI-generated answer, even if nothing on it is wrong.

Optimising your own assets lets you control the narrative. If you don’t publish honest comparison content, customers will piece it together from reviews, user content or competitors instead.

You're probably feeling this if:

• Plenty of traffic arrives but rarely turns into an enquiry or a sale
• Your blog is active but enquiries haven't moved
• The answer to a common question is buried three paragraphs into a page
• Some of your best-performing pages haven't been meaningfully updated in years, even though the information on them has moved on

Layer 4: Your entire online reputation

Nothing you do on your own website decides how you're described everywhere else and everywhere else is now part of how people decide to trust you. Reviews, directory listings and mentions elsewhere shape both what a customer thinks and what an AI tool says about you.

What this involves

• Keeping your Google Business listing accurate and active
• Understanding who's linking to your site and building genuine links from other credible, relevant websites, not just increasing the number
• Encouraging honest, recent reviews across the platforms your customers actually use, not just Google, that mention the specific service someone used rather than just how happy they were overall
• Making sure your business is described the same way everywhere, not slightly different details on every site

A note on review websites, citations and link building

Reviews matter for AI answers too, but not just the star rating and not just on Google. AI tools draw on reviews wherever they can find them. Dedicated review sites like Trustpilot are consistently among the most cited sources in AI-generated answers, alongside industry directories and whatever else your customers use. 

When someone asks an AI tool whether a business is good at something specific, it leans on what customers have actually written, not how many stars they left. A generic "great service, highly recommend" does very little here. If you offer several services and your reviews mostly talk about one of them, that's often what AI ends up confident recommending you for too, even if you're just as good at the rest.

Getting quoted in an article or building real credibility usually needs input from people outside marketing and it can't be squeezed into a content calendar. It works best as an ongoing effort with someone clearly responsible for it.

Link building matters most as a tie-breaker. Once your website is properly comparable to a competitor's on the things you can control, similar content, similar technical health, it's the off-page signals, like who links to you, that start to decide it. Think of it like a job reference. A recommendation from your old CEO carries more weight than one from the person who happened to sit next to you. It's the same principle with websites, a mention from a respected, well-known site tells Google and AI far more than a mention from somewhere nobody's heard of.

You're probably feeling this if:

• Competitors with a weaker website still consistently outrank you
• Your reviews are thin, outdated or inconsistent across platforms
• Your business listings still describe services you no longer offer
• Your reviews are generic and rarely mention the specific service someone used, especially if you offer more than one
• Similar competitors seem to have links from bigger, more relevant sites than you do

Layer 5: Reporting on what's working

This layer is the outcome rather than the work itself and it needs a light touch in how we report on it, because different platforms behave very differently from each other. Google's AI Overviews are a good example, they've already reshaped how many clicks even a first-place ranking gets and they don't behave anything like a normal search result.

Checking what ChatGPT or Gemini say about your business is useful for spotting trends, but it's a temperature check and not a scoreboard. Ask the same question twice and you can get two different answers, so you shouldn't use a handful of AI checks alone to prove a point either way.

Both Google and AI tools also tailor what they show based on things like your location and your own past searches or conversations. That means checking your own rankings from your own device, or asking ChatGPT about your industry from an account it already knows plenty about, doesn't necessarily show you what a genuine new customer sees. It's a natural thing to want to check yourself, but it's exactly why we track this with a consistent, neutral set of searches and prompts, rather than relying on what any one person happens to see on their own screen.

The measures we'd actually report on are your visibility in normal search results, how much of that traffic reaches the right pages and how many real enquiries or sales come through. AI visibility checks help us work out where to focus next, rather than being a number to chase.

Why "AI SEO" isn't a new discipline

You'll increasingly hear terms like Answer Engine Optimisation (AEO) or Generative Engine Optimisation (GEO), essentially being visible when someone asks ChatGPT for a recommendation instead of Googling it. But the work behind this isn't entirely new or mysterious. Many AI tools draw on the same signals, sources and high-quality content that already help pages perform well in traditional search.

Google's AI Overviews are generated straight from its own search index. Independent studies of ChatGPT and Perplexity have repeatedly found a real correlation between ranking on page one of Google and turning up in AI-generated answers. It's not a perfect one-to-one relationship, other things matter too, but ranking well remains one of the most reliable ways to end up in both. 

In reality, search is shifting from rankings to visibility rather than turning into a different discipline and a site that's easy to understand, has genuinely useful content and a good reputation elsewhere tends to get picked up by AI tools for the same reasons it does well on Google.

Where we focus first

All five areas matter from day one of working together. What changes from client to client is where we spend the first few months and that depends on what's currently holding you back most.

The biggest problem in search is often effort in the wrong place. More content won't fix a site search engines can't read, and more links won’t help weak pages. Getting the order wrong is one of the costliest mistakes we see.

This is how Bright Labs works as an SEO agency in Melbourne. All five areas are in play from day one, weighted to whatever's holding you back most, with the different types of work happening alongside each other rather than one after another. If a piece of this sits better with your own team, we'll say so upfront rather than leaving a gap halfway through.

If you're weighing up your search and marketing plans for the year ahead, it's worth a conversation before anything gets locked in.