GEO & AEO (AI Search Optimisation)

‍

Every company will care about how AI talks about their brand.

Your buyers are already asking AI which companies to consider, how products compare, and what solution fits a specific problem. Those answers can shape the shortlist before someone reaches your website.

We help B2B companies measure what AI says about them, understand where competitors are more visible, and improve the content and signals behind those answers to increase visibility. We start with a focused baseline audit using around 25 questions that matter to your business.
‍

What do GEO and AEO mean now?

The terminology is still settling. Answer Engine Optimisation, or AEO, and Generative Engine Optimisation, or GEO, often describe overlapping work. On this page, we use the plainer term "AI search optimisation".

The acronym matters less than the outcome. Your company should appear when buyers ask relevant questions, be described accurately, and be recommended for reasons you can support. The sources behind the answer matter too.

In a Webflow survey of more than 100 marketing leaders and 300 practitioners, 93% of leaders said AEO would be important to company success within two years. Only 25% of practitioners said they fully understood it. [cite:Webflow survey|https://webflow.com/blog/marketers-ai-adoption-wave|Webflow marketer survey, 2025. Shows strong AEO interest and a large understanding gap.]

That gap is where practical measurement helps.

‍

What are we trying to improve?

The work has three main goals.

Close obvious citation gaps

These are questions where your company should appear but does not. A competitor may be cited for a product category, use case, or buying criterion that your website already covers. The first job is to understand why the connection is missing and strengthen the relevant source material.
‍

Become a credible recommendation

A mention is not always a win. The answer may describe your company inaccurately, associate it with the wrong category, or recommend a competitor because your proof is weak or hard to find. We look at the context of the mention as well as its frequency.
‍

Claim unclaimed spaces

Some valuable questions have no consistent winner yet. They may concern an emerging problem, a niche use case, or a buying question the market has not answered well. These spaces can become strong content opportunities when they match a real business priority.
‍

Does traditional SEO still matter?

Yes. AI search is another discovery layer built partly on the same web that powers traditional search.

Google says the normal SEO fundamentals still apply to AI Overviews and AI Mode. Pages need to be crawlable and indexable. Their content needs to be useful and connected through clear internal structure. Google also states that no special AI markup is required for inclusion. [cite:Google Search Central|https://developers.google.com/search/docs/appearance/ai-features|Google guidance, updated 2025. Existing SEO fundamentals remain relevant to AI features.]

The difference is in what we measure. A traditional rank tells you where one page appears for one query. AI answers can combine searches and sources. They can mention several brands, change their wording, and produce different results across platforms.

SEO remains the foundation, but no longer shows the whole visibility picture.
‍

How does the initial audit work?

Before an ongoing engagement, we define around 25 prompts. They reflect your products and customer problems, along with the buying criteria and strategic priorities that matter most.

We run them across ChatGPT, Google AI Overviews, and Perplexity. The three platforms do not retrieve and assemble information in the same way, so looking at them separately matters.

The audit shows data like:

• How often your company appears.
• How you compare with selected competitors.
• Whether performance changes across the three platforms.
• Which business themes are strong or weak.
• Where the clearest content, authority, or accuracy gaps appear.

We then walk through the findings with you. The goal is not to turn one score into a verdict. We establish a useful baseline, explain what the metrics mean, and decide whether the opportunity is worth pursuing.

The audit is free and does not require an ongoing agreement.
‍

What happens after the audit?

An ongoing programme expands the baseline into a controlled measurement and implementation loop.

1. We build the prompt map with you

We develop around 100 questions organised by strategic business theme. Available search data informs the set. So do customer language, sales questions, internal knowledge, market priorities, and your team's input.

The prompts cover different stages of the buying journey. We include problems and solutions, use cases, technical requirements, ROI, education, comparisons, and alternatives.

For a B2B fintech platform, one theme might cover digital client onboarding and another regulatory reporting. Portfolio operations and vendor comparison may need their own themes. This shows whether the brand is visible in one part of the business but absent in another.

2. We track the questions every day

The agreed prompt set runs across ChatGPT, Google AI Overviews, and Perplexity every day. With around 100 prompts and three platforms, that produces about 300 answer checks per day and 9,000 per month.

Repeated tracking matters because generated answers vary. A single screenshot can reveal a problem, but it cannot show a trend.

3. You get a custom dashboard and monthly report

The dashboard shows overall visibility and performance by platform. You can also inspect competitor rankings, business themes, and individual prompts. Each month, we publish a report showing what moved, what stayed flat, what we completed, and what to prioritise next.

4. We improve the signals behind the answers

We typically work on a monthly retainer that combines ongoing monitoring and reporting with a fixed number of implementation hours. We use those hours to address the gaps we find, rather than handing over a report and leaving the client with a list of tasks.

5. We measure again

The next reporting cycle shows whether the tracked answers reflect the work. Some changes work. Others do not. The loop lets us learn and adjust.

One AI answer is an anecdote. Repeated answers across a fixed prompt set become a useful signal.
‍

What do we measure?

We look at the headline numbers and the answers behind them.

The most important metrics are:

• Visibility score: The percentage of brand-mentioning answers in which your company appears.
• Share of voice: Your share of all brand mentions across the tracked answers.
• Average position: How early your company tends to be mentioned when it appears.
• Citation share: How often your own pages or assets appear among the cited sources.
• Accuracy and recommendation quality: Whether the answer understands your company, connects it with the right use cases, and gives buyers credible reasons to consider it.

We break these metrics down by platform, business theme, prompt, and competitor. That makes the results more useful than one overall score. A company can be visible for its established product while almost absent from a new strategic category.
‍

How reliable are AI visibility metrics?

They are directional, not absolute.

Generated answers can change with the model, date, location, source index, and phrasing of the question. No single universal AI rank behaves like a traditional search position.

We control the prompt set, platforms, frequency, and reporting method. Daily tracking reduces the influence of one unusual answer and makes sustained movement easier to see.

We also keep the raw answer context. A rising visibility score means little if the answer describes the company inaccurately or for the wrong reason.
‍

How long does it take to see results?

In our work, AI answers often respond much faster than traditional SEO rankings. We can sometimes see the first effect within days after AI systems pick up an updated page or source.

That speed is one reason daily tracking matters. One changed answer can be noise. When the same improvement appears across repeated runs and more than one platform, it starts to look like a real result.

Not every gap moves at the same speed. Clearer on-site content, corrected company information, and stronger proof can have a quick effect. Broader changes in authority, recommendation patterns, and third-party coverage still take longer.

We do not promise a fixed timeline, but the feedback loop moves much faster than traditional SEO work.
‍

What do you actually optimise?

We start with the technical foundations and content systems on the website, then work outwards to the sources shaping AI answers.

On-site optimisation

• Crawlability and indexing: We review URL structure, canonicals, redirects, sitemaps, broken links, and page-level indexing signals.
• Site architecture and internal linking: We review and optimise key pages and links so the relationships between topics, services, and expertise are easier to follow.
• Agent-readable content: We maintai llms.txt and create text or Markdown versions of important pages so agents can access the core information without working through the full visual layer of the site.
• Structured data systems: We implement and maintain schema across page templates and CMS collections for the company, its services, articles, people, and other relevant content.
• FAQ infrastructure: We build reusable FAQ modules that connect visible questions and answers with matching schema, then support the client in shaping answers that address gaps revealed by customer research and prompt tracking.
• Entity clarity: We help structure facts supplied by the client's team and present them consistently across relevant pages.
• Page and content gaps: We identify where a new or improved service, use-case, comparison, or educational page could close a visibility gap. The client provides the subject matter; we help shape the structure and implement it.
• Proof and maintenance: We help surface the client's cases, credentials, examples, and original evidence, then keep their presentation current as the market changes.

Off-site optimisation

• Citation and source analysis: We identify which external sources appear in tracked answers, where competitors gain visibility, and which gaps are worth pursuing.
• Profiles and industry sources: We identify relevant profiles, directories, partner pages, and industry platforms, then help strengthen the company information they contain.
• Reviews and third-party proof: We identify gaps in review coverage and help surface credible customer evidence on independent platforms.
• Editorial presence: We identify opportunities for expert contributions, partner content, data, or stories that can earn credible mentions and citations, then support the client in developing them.

The client owns the subject matter and supplies the product knowledge. Its customer insight and strategic priorities guide the content. We identify the gaps and help structure the material, then implement it on the website.

The tracking tells us which of these areas deserves attention first. We make the changes, then measure how the answers respond.
‍

What makes Oimachi's approach different?

We do not stop at monitoring or recommendations. We work proactively from the data and implement improvements directly on the website.

Because Oimachi works across positioning, content, website structure, Webflow development, and technical implementation, we can act on what the tracking reveals. That might mean rewriting a product page, creating missing comparison content, improving the CMS structure, adding proof, or correcting a technical issue.

Your team brings knowledge of your customers, sales conversations, and strategic priorities. We use that context to decide which questions to track and what to change. Then we make the changes and measure whether the answers improve.
‍

How is the ongoing work priced?

We structure ongoing work as a monthly retainer based on how much optimisation capacity the company wants. We recommend a minimum of 15 hours per month. That gives us enough time to act on the findings instead of limiting the work to reporting.

Some companies start at that level. Others want to move faster and invest more.

Every ongoing engagement includes monitoring, access to the custom dashboard, and a monthly report. The main variable is the monthly number of hours available for content, technical, and off-site improvements.
‍

Start with a baseline

You do not need a complete AI search strategy before you begin.

Start with the questions your customers ask. We will show you where your company appears, where competitors are ahead, and which gaps are worth acting on before you commit to ongoing work.