Our team uses machine learning and mechanistic interpretability to influence how AI thinks about brands, products and services. We cover the dominant AI ecosystems including Google, OpenAI and Anthropic.
This includes AI search, Google's AI Mode, AI Overviews and AI assistants and agents such as Gemini, ChatGPT and Claude. Reserve your time slot for a conference call with our senior AI Influencers to discuss your project:
Engaged by global brands.
How we work
01
Map Connections
We nurture a strong culture of testing and measuring. We like to know what works, what doesn’t work, and most importantly, we like to know why.
02
Find Connection Strength
We innovate all the time. It’s in our DNA. When working together with your team we’re very likely to come up with something that’s never been done before.
03
Selection Rate Optimization
We see ourselves as an extension of your team and take great care to ensure that you understand our work. Our best campaigns are based on strong collaboration.
Content Optimizer is a content optimization engine that aligns your page content with AI model preferences, making your brand more likely to be selected, cited, and recommended in AI search and chat. It optimizes your pages for the way AI assistants evaluate, compare, and choose sources.
Optimizes for the decision, not the ranking
Traditional SEO optimizes for where you sit on a results page. Content Optimizer optimizes for something different: whether an AI grounding system would choose to quote you.
The mechanism is direct. When an AI assistant answers a question, it compares competing sources and selects the most quotable passage. Content Optimizer simulates that exact decision with an AI ranker — then iteratively rewrites your page, or the snippet an AI would lift from it, until the ranker prefers your content over the competition.
It isn’t a checklist or a static score. It’s a measured contest, run round after round, until your page is the one the model picks.
Client Success: OWAYO
| Metric | Apr 15, 2026 | May 31, 2026 | Percentage Points Up | % Increase |
| Share of Voice | 2.18% | 3.87% | +1.69 | +77.52% |
| Mention Share | 2.06% | 4.37% | +2.31 | +112.14% |
| Citation Share | 2.30% | 3.38% | +1.08 | +46.96% |
At the start of the AI Visibility campaign OWAYO wasn’t being recommended in AI assistant chat sessions and AI Mode for audiences in the USA.
Using our bayesian content optimizer we found that the brand was overly EU-centric causing models to withhold recommendations for the audiences in the USA.
An on-site optimisation followed by a 6 month off-site brand alignment campaign resulted in OWAYO’s AI visibility by up to 90% per entity and 2% global uplift for all targeted entities.
What an engagement looks like
Every engagement runs the same five stages. The figures below are what each stage produced on the OWAYO campaign, so the shape is concrete rather than abstract.
We list the entities that decide your category, connect each one to the page that should own it, and take a tracking baseline before anything changes. For OWAYO that was 192 entities and 83,827 modelled associations between products, terms and pages, with 31 entities held back as the visibility tracking set every later number is reported against.
We line the AI visibility picture up against your existing organic footprint and look for the disagreements: entities you already rank for in traditional search that models do not associate with your brand.
We find the sources AI systems actually retrieve for your topics, so outreach targets the pages models read rather than the pages that look authoritative. The OWAYO run covered 181,686 citations across 12,610 domains.
We probe the models directly and record what they hold true about your brand, one finding at a time, then rewrite content to correct the weak, missing or wrong associations. OWAYO’s ledger reached 963 findings across 113 Content Optimizer sessions.
On-site changes affect what a model retrieves from you. Off-site signals affect what it associates with you when your site is not the source. The two run in parallel, and the off-site half is scoped over six months because brand signals take that long to settle.
Reporting runs against the stage-one baseline on three numbers: mention share, citation share, and average position inside the answer.
DEJAN’s methodology transcends traditional AI SEO, diving into the core mechanics of LLMs to provide actionable intelligence for AI visibility. We understand that AI models, while appearing intelligent, operate on statistical probabilities and learned associations. Our tools are designed to surface these underlying mechanisms, providing clarity on how and why models make certain decisions.
Deep Understanding
We move beyond surface-level metrics to analyze log probabilities, token flow, and decision-making junctions within LLMs, identifying precise points for optimization.
Actionable
Our insights translate directly into actionable strategies for content creation, internal linking, and even influencing user prompting patterns, ensuring your brand’s message resonates effectively with AI.
Testing.
We nurture a strong culture of testing and measuring. We like to know what works, what doesn’t work, and most importantly, we like to know why.
Innovation.
We innovate all the time. It’s in our DNA. When working together with your team we’re very likely to come up with something that’s never been done before.
Collaboration.
We see ourselves as an extension of your team and take great care to ensure that you understand our work. Our best campaigns are based on strong collaboration.
Australia
DEJAN has run search campaigns from Australia since 2008, more than a thousand of them. The company is registered in Queensland and the team works remotely from Brisbane, Melbourne, Sydney and Perth. Australian brands on the client list include Virgin Australia, Xero, Sportsbet, Australia Post, Beyond Bank, Ubank, Compare the Market, Healthengine, Lendi and Griffith University.
Market matters more in AI search than it did in ranked search. The sources a model retrieves to answer an Australian buyer are not the sources it retrieves for the same question in the United States, so a brand can be named in one market and absent in the other. Every measurement we take is tied to a market for that reason. The OWAYO campaign above was tracked in United States answers; the same instruments run against Australian answers for Australian clients.
The question we answer for an Australian brand is whether Google AI Mode, Gemini, ChatGPT and Claude name you when a buyer in Australia asks for a recommendation in your category. We measure that first and give you the baseline before any optimisation starts, so the campaign is judged against a number you saw on day one.
Meet our core team
We’re an all-senior team with experience in a wide range of projects and industries.
Mike Jolly
Director of Strategy
Blake Walsh
SEO
Giordano Chng
SEO
Liam Buttery
SEO
AI SEO
Martin Reed
Technical SEO
Bianca Hall
Public Relations
Alex Petrovic
SEO
Danielle White
Operations
Milos Dosen
CFO
Josip Ivanovic
Developer
Nemek Nowaczyk
PPC
Dragan Grubacki
Technical SEO
Finn Arrowsmith
Outreach
We were given our very own bespoke internal link recommendation engine that leverages world-class language models and data science. It’s one thing to theorize about the potential of machine learning in SEO, but it’s entirely another to witness it first-hand. It changed my perspective on what’s possible in enterprise SEO.
Scott Schulfer
Senior SEO Manager
Zendesk
Featured In
Dan Petrovic, an academic and consultant on SEO and generative AI, said Google’s size, expertise and massive trove of search data gave it a massive advantage, but that Gemini 3 Pro would probably be a more expensive model to run.
Dan Petrovic made a super write up around Chrome’s latest embedding model with all the juicy details on his blog. Great read. Jason Mayes, Google Web AI
Common questions
AI SEO at DEJAN starts at $5,000 per month. Scope sets the number: how many entities we track, how many markets and languages, and how much content production sits inside the retainer. We quote after a call rather than from a price list.
OWAYO’s share of AI brand mentions moved in the first full month of tracking: 1.5% in March 2026, 3.8% in April, 5.2% in May, 6.5% in June and 8.0% in July. Average position inside an answer went from 5.6 to 2.9 across the same period. On-site work shows first, off-site brand signals take longer, and a full campaign is scoped over six months. The numbers and the method are in the OWAYO case study.
DEJAN SEO PTY LTD is an Australian company, ABN 77 151 340 420, registered in Queensland and running search campaigns since 2008. The team works remotely from Brisbane, Melbourne, Sydney and Perth. We keep no office for visitors and publish no phone number. Meetings are video calls, which is how we also work with clients in Europe and the United States. Australian and global brands sit on the same client list, including Virgin Australia, Atlassian, Xero, Zendesk and OWAYO.
Yes. AI answers are assembled from retrieved web pages, so the pages a model can find and trust decide which brands it names. OWAYO went from 1.5% to 8.0% of every brand named in US AI answers about custom sportswear in four months, measured across 6,776 tracked answers. Rankings no longer describe the outcome, so we track mention share, citation share and average position inside the answer.
They do the mechanical work well. Crawling, log analysis, schema generation, entity mapping, bulk drafting and internal link modelling all run faster and more completely under an agent than under a person. They are unreliable on judgment: which entities a brand should own, why a model associates a competitor with your category, what to publish and what to delete. We build the agents we run, and a senior strategist signs off on what they produce.
We publish a directory of the people doing this work at Best AI SEO Agencies. It profiles individuals instead of ranking logos, and DEJAN appears as the curator rather than as an entry, because most “best of” lists are written by the agency sitting at the top of them. When you are comparing agencies, ask each one to show measured movement in AI answers and to name the model, the market and how many answers they measured across.
Book a conference call with our senior strategy team to discuss your project in detail.
