Rankings on Google, accurate presence in ChatGPT and Gemini answers: two kinds of search, one plan.
We treat SEO and GEO not as separate jobs but as two sides of the same visibility problem. Technical SEO, content, local SEO and a clean multilingual setup get your site found on Google. GEO is about your brand being mentioned accurately, currently and favourably in the answers given by ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews. Both rest on the same foundation.
From Istanbul we work on sites serving Türkiye, the UK and Uzbekistan, measuring Turkish and English queries separately. Every engagement starts with an audit: Search Console data, crawlability, page speed, structured data and where your brand stands in AI answers today. From there we build a prioritised task list and implement it alongside your team or entirely ourselves.
- 01Companies with a strong offer that still trail their competitors on Google.
- 02Brands that rank well on Google but never come up in AI answers.
- 03Businesses that have noticed AI recommending a competitor instead of them.
- 04Businesses with branches in Istanbul or several cities that need to be found in local searches.
- 05Brands that need the Turkish and English versions of their site served correctly in several markets.
- 06Teams about to migrate, redesign or replatform who want to keep the traffic they already have.
What SEO and GEO are, and how they differ
SEO, search engine optimisation, is the technical, content and authority work that gets a site ranking for relevant queries on search engines such as Google. The user sees a list, clicks a few results and makes up their mind on the websites themselves. The yardsticks are well known: impressions, rankings, clicks and ultimately conversions, all of which can be tracked in Google Search Console and your analytics.
GEO, generative engine optimisation, is the work of getting a brand into the answers produced by AI engines such as ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews. Often there is no list at all: a single paragraph names three to five brands and the user rarely looks beyond it. The yardsticks change too: instead of rankings you track mentions, citations and whether the information given is correct.
We handle the two together for a simple reason. Many AI engines search the live web while composing an answer, and they favour pages that are well structured, crawlable and trustworthy. A weak SEO foundation weakens GEO as well, while an SEO plan that ignores GEO leaves empty the very place where people increasingly go to ask their questions.
How classic SEO works: technical, content, local and international
Technical SEO is the foundation. Using Search Console and crawling tools we check whether the site is being crawled, which pages are indexed, redirect chains, duplicate content, Core Web Vitals and the mobile experience. Whether you run Next.js, WordPress or a custom stack, we write issues up clearly enough for your developers to act on, or fix them ourselves if you prefer.
On content, we group keyword research by search intent and aim for one strong page per intent. Merging pages that compete with each other, rewriting titles and descriptions and tidying internal links often moves the needle faster than publishing anything new. For local SEO we cover Google Business Profile, branch pages, consistent name, address and phone details, and review management.
On sites serving several markets, getting hreflang tags, language and country structure, currency and localised content wrong means the Turkish page shows up in the UK and the English page in Türkiye. So we plan multilingual setups for Türkiye, the UK and Uzbekistan from day one. For site migrations, we map every old URL to its redirect before the move, so existing traffic survives it.
AI search visibility: how do you show up in ChatGPT?
The first step is building the question set. We map the questions your customers really ask, along the lines of “which one is most reliable”, “what’s the difference between these two” or “who does this in Istanbul”, in Turkish and English, and put them to several engines at regular intervals. For every answer we record whether the brand is mentioned, which competitors appear alongside it and which sources are cited.
Then we answer the question “why aren’t we there?” and work on three tracks. The first is entity building: the brand’s name, category, founders, products and location need to appear the same way on your site, in directories, on social profiles and in reference sources, so models recognise it as one clear entity. The second is the technical layer: schema.org structured data, access for AI crawlers, an llms.txt file that summarises the site, and pages that answer questions directly and are easy to quote.
The third is source work: getting the brand represented accurately in the lists, comparison sites and trusted publications your category relies on. We don’t use shortcuts that try to game AI answers; the only thing that survives model updates is real, consistent information.
Timelines, measurement and the most common mistakes
Technical fixes and page updates can ship quickly. Google usually reflects them within weeks, and engines that search the live web, such as Perplexity and AI Overviews, respond at a similar pace. Lasting rankings for competitive keywords, and changes in answers drawn from a model’s training data, build up over months. That is why we run SEO and GEO as a programme measured in quarters, not as a campaign.
We measure from two sources: Search Console and analytics for rankings, clicks and conversions, and a monthly re-run of the same question set for mentions, citations and the accuracy of each answer. AI answers are volatile by nature, so we read the trend across repeated measurements rather than any single screenshot.
The mistakes we see most: judging traffic by visitor numbers rather than revenue, publishing dozens of pages that compete for the same intent, forgetting redirects during a migration, leaving hreflang half-finished, writing separate content for AI that no human reads, and measuring one engine then assuming the rest behave the same. Showing up in ChatGPT with the wrong price or a closed branch can do more harm than not showing up at all.
- 01Technical SEO audit and fix list
- 02Keyword and question map in Turkish and English
- 03Content plan and on-page optimisation
- 04Local SEO and Google Business Profile clean-up
- 05Structured data, llms.txt and entity building
- 06Monthly Search Console and AI visibility report
Two kinds of search, one plan
We don’t split Google rankings and visibility in AI answers across separate teams and separate reports. One audit, one content plan and one set of technical fixes serve both, so your budget is never spent twice on the same work.
A team that ships AI products
We work with large language models every day in Rocketly CRM, ad10 and Kaynak. We learned how models pick sources, summarise information and get things wrong by building products on them, at the level of code.
A media and sales lens
With more than 20 years in media and sales leadership, we tie GEO to commercial goals. We work out which questions sit close to a purchase and which are purely informational, and prioritise accordingly.
Measured, with no shortcuts
Every month we report Search Console data and recorded AI answers side by side. We don’t use tricks that try to game rankings or AI answers; lasting visibility comes from real, consistent information living in trusted places.
Audit
We review Search Console, crawling, indexing, speed and structured data, and at the same time measure where the brand stands in AI answers today.
Map
We map the keywords customers search on Google and the questions they ask AI, in Turkish and English, ranked by how close they sit to a purchase.
Implement
We ship technical fixes, content updates, schema markup and consistent brand information across external sources, strictly in order of priority.
Measure
Every month we report rankings, clicks and conversions from Search Console, and mentions and citations from AI engines using the same question set.
Scope comes down to a few factors: the size and stack of the site, how many languages and markets you target, how many locations need local SEO, the question set and AI engines we track, and who implements the changes. If your team handles technical and content work, we focus on audit, prioritisation and direction; if you want us to implement as well, the scope grows. The initial audit is priced as a one-off project, followed by a monthly programme.
Let’s talk about thisSEO is the work of getting a site to rank for relevant queries on search engines. It rests on three pillars: technical foundations that let search engines crawl and understand the site, content that genuinely answers the searcher’s intent, and links and mentions from other trustworthy sites. Results are tracked in Search Console as impressions, clicks and rankings.
GEO is the work of getting a brand into the answers of AI engines such as ChatGPT, Gemini, Perplexity and Google AI Overviews. SEO aims to rank a page high in search results; GEO aims to have the brand mentioned, accurately, inside a single answer the AI has put together. Measurement differs too: instead of rankings, you track mentions and citations.
Technical SEO covers crawlability, indexing, page speed and Core Web Vitals, mobile usability, redirects, duplicate content and canonicals, the sitemap and robots.txt, structured data and, on multilingual sites, hreflang. In short, it is the layer that makes sure search engines can reach your content and understand it correctly, however good that content is.
The usual reasons: your brand doesn’t appear enough in the sources your category relies on, the information on your site isn’t clear enough for a model to use, your brand name is written inconsistently across the web, or competitors have content that answers the question more directly. Sometimes the model is simply working from old data. Diagnosis exists to find out which applies to you.
Start by being clear about which questions you want to appear for. Then answer those questions plainly and quotably on your site, set up structured data properly, make sure AI crawlers can actually reach your pages and get your brand represented consistently in trusted external sources. It works when all of that runs alongside regular measurement.
llms.txt is a plain-text file placed at the root of a site that tells AI tools what the site is, which pages matter most and what the content covers. It isn’t yet a standard every engine officially supports, and it won’t create visibility on its own. But it is cheap to produce and, alongside structured data and consistent information, makes a site easier for machines to read.
AI Overviews are AI-generated summaries that Google shows above the regular results for some searches. Because people often get their answer right there, clicks to websites can drop, while the sites cited inside the summary gain visibility. That makes it increasingly important for your content to be clear and trustworthy enough to be used as a source.
The effect of technical fixes and updates to existing pages usually shows up in Search Console within a few weeks. Lasting rankings for competitive keywords tend to take a few months, and longer on a new domain. In AI answers, engines that search the live web move quickly while those based on training data move more slowly. Expect a first report in weeks and a meaningful trend in months.
Steer clear of anyone promising guaranteed rankings, staying vague about what they actually do or setting up accounts in their own name. Look for a team that starts with an audit, writes down its priorities with reasons, leaves Search Console access with you and reports on conversions rather than raw traffic. Today, the ability to measure AI search visibility belongs on that list too.
I usually reply within 24 hours.