SEO and AI search. Brisbane, working Australia wide.

When Someone Asks AI For A Recommendation, Does It Say Your Name?

We are an SEO agency for Google and AI search. Most brands have never checked. The ones that have usually find the same thing: the AI knows they exist, and still names someone else. We find out why, and we fix it.

A real example

This Is A Real AI Answer About A Real Australian Brand. It’s Wrong.

In August we ran an AI visibility baseline for Toomi, an Australian womenswear label. We asked the engines what sizes its dresses come in.

The AI didn’t invent the answer. It read a collection page title, took the number as fact, and started republishing it to everyone who asked. One inaccurate line on one page became the answer the internet gives about that brand.

We wrote that title. The mistake was ours, and we found it in public, in an answer we could not edit. That is exactly what this service exists to catch, and it is why we start by looking rather than by selling you a package.

Google AI Overview, recorded 10 Aug 2026

“Toomi is an Australian maximalist fashion label known for its vibrant, hand-illustrated prints, 100% cotton fabrics, and dopamine-dressing style. Their dresses come in extended sizes ranging from AU 8 to 26.”
AI answers change between sessions, so this is what came back on the day, not a fixed result. Those dresses run AU 10 to 22.
0 of 11 discovery questions that named the brand

Source: Toomi AI SEO baseline, 10 August 2026. Asked about by name, both engines described the brand accurately.

2022 age of an article still shaping answers

Source: same baseline. A September 2022 Brisbane roundup was still being cited in 2026.

8 to 26 the size range Google published as fact

Source: same baseline. Those dresses run 10 to 22, and the false range came from a page title we wrote.

The actual problem

The Gap Isn’t Your Reputation. It’s What You Give It To Quote.

When we audit a business for AI search, we find the same thing almost every time. The AI already knows the brand exists. It just does not name it when someone asks for a recommendation.

That gap is not about how good you are. It is about whether your site gives an AI something it can quote. Most give it nothing, because they were built to rank on Google, which is a different job. Ranking first and being named in the answer are two different systems, and doing well at one does not carry over to the other.

You Need This If Any Of These Is True

  • You have never checked what AI says about your category. If you have not looked, you do not know who is being named instead of you.
  • Your site has no questions and answers on it. AI quotes sentences. Product and brand pages do not give it any.
  • Your details are inconsistent across the web. Different numbers, hours or descriptions teach the AI not to trust you.
  • You are relying on your Google ranking. Ranking first and being named in the answer are two different systems.

The mechanism

AI Doesn’t Read Your Website To Answer “Who’s Good At This?”

This is the part most agencies get wrong, and it changes what the work should be.

When we traced where the answers actually came from, not one of them was assembled from a brand’s own website. They were built from editorial roundups, magazine lists and Reddit threads. One of the sources still shaping answers in 2026 was a Brisbane article published in September 2022.

So an agency that only tidies up your website is solving half the problem, slowly. Your site is what makes you quotable once you are found. Being found in the first place happens somewhere else, in places with long memories. We work on both, in that order, and we will tell you which one is holding you back before you spend anything.

Source: Toomi AI SEO baseline, 10 August 2026. Recurring sources included Urban List, Vogue AU, frankie, Peppermint and Reddit.

Straight answer

If You’re On Shopify, Half This Category Is Selling You Something You Already Have

There is a file that tells AI systems how to read your site. It has become the headline deliverable in a lot of AI search pitches this year.

Every Shopify store already has one. It is generated for you, it costs nothing, and there is no app to install. The same goes for the machine readable product feed and the agent endpoints underneath it. If someone quotes you to build those, they are quoting you for work that is already done.

We would rather lose the line item than bill you for it. What Shopify does not give you is the part that says who you are rather than how to buy from you, and that is the part that gets you named. So that is what we work on.

What We Actually Do Instead

Fix the facts about your business so every version of them agrees. Build the pages an AI can quote a sentence out of. Get you into the third party sources the engines actually read. Then check, every month, whether the answers changed.

Roadblocks

Five Reasons Brands Put This Off

Every one of these comes up on a first call. Here’s what we do about each.

“Nobody’s searching for us in ChatGPT yet, this feels early.”

What we do

It is early, and that is the argument for doing it now rather than against it. The sources shaping answers today are years old. A 2022 article was still feeding answers we pulled this August. Whatever gets written about you now is what the engines will be quoting in three years, and the brands already in those sources will not be easy to displace later.

“We already pay for SEO. Isn’t this the same thing?”

What we do

No, and your SEO agency is not necessarily doing it. Ranking first gets you a link on a page of links. Being named in the answer means the AI repeats your name to someone who never sees a page of links at all. The technical work overlaps a little. The content work barely overlaps at all.

“How would we even know if it worked?”

What we do

We track the same set of real buying questions every month and record what each engine says. Not a score out of a hundred. The actual answers, with the dates, so you can see them change. There is a whole section below on how we handle the fact that AI answers are never identical twice.

“We don’t have anyone to write more content.”

What we do

Neither do most of our clients. We write it. The plans include the pieces, written to answer the questions your buyers actually ask, because those are the ones an AI can lift a sentence out of.

“We got quoted a quarter of your price for this.”

What we do

You probably did, and some of that quote is for the Shopify files we just told you are free. Ask whoever quoted you two questions: what does the AI say about my brand today, and where did that answer come from. If they cannot answer both, they are selling you a checklist. We will answer both before you pay us anything.

The work

What’s Actually Included

No mystery retainer. This is the work.

01

Where you stand in AI answers today

We ask the engines the questions your buyers ask and record exactly what comes back. Rechecked every month on the ongoing plans, so you can see whether anything moved.

02

Your website fixed and your five biggest pages rewritten

The pages that carry the most weight, rewritten so an AI can quote them, and the facts underneath them corrected so every version agrees.

03

The other pages AI quotes you from, cleaned up

The ones you did not know were being read. One new one a month on the middle plan, two a month on the top one.

04

New content written to get you quoted

Written to answer real buying questions. Two pieces a month, or four on the top plan. Not included in the one-off audit.

05

A report in plain English

What the answers said, what changed, what we did. Quarterly or monthly depending on the plan. No jargon, and no dashboard you have to learn.

Every plan starts with the audit, because we will not sell you fixes before we know what is broken.

The honest bit

AI Answers Are Never Identical Twice. Here’s How We Deal With That.

Ask ChatGPT the same question twice and you will get two different answers. Any agency that hands you a single visibility score is hiding that from you.

We do it the way you would test anything unreliable: the same questions, on a schedule, recorded every time. One absence tells you nothing. The same absence three months running is real, and so is the month your name starts appearing. You get the actual answers with their dates, not a number we made up to summarise them.

It also means we will tell you when a result is directional rather than proven. That is a slower kind of reporting and it is the only kind that is true.

The other thing you’re buying

Small, Aussie, And Straight With You

No offshore outsourcing, no account manager relaying messages to someone you never meet. Fixed prices, no lock in contracts, and a person who answers you. We are a small Brisbane team and that is the point: you get the people doing the work, not a summary of them.

What it’s like working with us

We have not been doing AI search long enough for anyone to have a five year story about it. This is a client talking about what we are like to work with, which is the part that does not change between services.

“I’ve worked with Jess and the OLS team for over 3 years, and I’m continually impressed by their ability to adapt to new trends, deliver engaging content, and improve performance month-on-month!”

Kirsty Visman, Superdream

Objections

Questions We Get Asked

What is AI search optimisation, and how is it different from SEO?

SEO gets you ranked on a page of links. This gets you named inside the answer, when someone asks ChatGPT or Google's AI for a recommendation and never sees the links at all. Some of the technical groundwork overlaps. The content work mostly does not.

Do I still need normal SEO as well?

Yes. Google search is still where most buying starts and it is not going away this year. Think of this as the second front, not the replacement. If your basic SEO is broken we will tell you, because fixing that usually helps both.

Do you just write an llms.txt file?

No, and if you are on Shopify you already have one for free. So does every other Shopify store, which is why it cannot be what makes an AI recommend you. The work that matters is making your facts consistent, giving the engines quotable sentences, and getting you into the third party sources they actually read.

How do you measure this when the answers keep changing?

Same questions, same schedule, recorded every time, with the dates. One absence is noise. The same absence three months running is a finding. We will tell you which one you are looking at.

How long before I see a difference?

The fixes to your own site land inside the first month. Getting named in answers is slower, because it depends on third party sources the engines trust, and those have long memories. Anyone promising you a date is guessing.

What does it cost?

Every plan starts with an audit, then runs monthly. We will quote you on the call once we have seen what shape your site is in. If you want a rough sense before that, ask us on the call and we will tell you straight rather than making you sit through a proposal.

Will this work for my industry?

It works wherever people ask AI for a recommendation, which is now most categories. Where it works best is anywhere buyers ask comparison questions before they buy. Where it is weakest is very local, very transactional services, and we will say so rather than take the work.

What do you need from my team?

Access to your site and your store, and one person who can confirm facts about the business when we find inconsistencies. That is genuinely it. The writing and the fixes are ours.

Find Out What AI Says About You

One call. We’ll run your brand against the questions your buyers actually ask, tell you what came back, and say whether there is a problem worth fixing.

What Happens On The Call

  • We ask the engines the questions your buyers actually ask
  • You hear the raw answers back, with the dates
  • If there is a problem worth fixing we say what it is, and if there isn’t we say that too