Reference material, not sales material
These guides exist so that your own team can do the work without us. Every number carries its source and its access date; no ratio is published without one, and nothing here is gated behind a form.
Which page should you start from, given your question?
If your question is why AI assistants never quote you, start with the AEO guide. If it is how to become a cited source in generative engines, start with GEO. If your site is slow, start with Core Web Vitals. If you are deciding where JSON-LD goes, start with Schema.org. For one term, go to the glossary.
The five resources here do not repeat one another; each describes a different layer of search visibility. The diagram below shows how the four guides relate. At the bottom sits the technical foundation that lets a machine load, read and measure your page — Core Web Vitals and Schema.org. Above it sits SEO, which aims to rank in the results page. Above that sits AEO, which aims to be quoted in the answer box. At the top sits GEO, which aims to be cited as a source inside generated answers. No upper layer holds for long while the layer beneath it is missing.
| The question you have | Where to start | What you will find there |
|---|---|---|
| Why do AI assistants never quote my site? | What is AEO? | The criteria answer engines use when they pick a passage, the length and position of a quotable passage, and how the work is measured. |
| How does my brand become a cited source in ChatGPT or Perplexity? | What is GEO? | The decision on AI crawler access, the correct size for an llms.txt file, entity clarity and how visibility is sampled. |
| My site is slow — which metric should I look at? | Core Web Vitals guide | Definitions and published thresholds for LCP, INP and CLS, why field data differs from lab data, and the order in which fixes are made. |
| Where do I put JSON-LD, and which type do I use? | Schema.org guide | A type-selection decision table for corporate sites, the logic of @id and the entity graph, validation tools and frequent mistakes. |
| What exactly does this term mean? | Glossary | Short, self-contained definitions across search, performance, technical infrastructure, structured data and artificial intelligence. |
| What is the measured state of my own site today? | Digital asset audit | A current-state report measured on five axes: AEO answerability, crawler access, Core Web Vitals, schema validity and WCAG 2.1 AA. |
These pages are not a purchase path. If you want to see who does which piece of work instead, the scope matrix of eleven services and how we work are the right addresses.
In which order is it most efficient to read these?
The suggested order has five steps: AEO first, then GEO, then Core Web Vitals, then Schema.org, and the glossary last. It runs from question to foundation rather than easy to hard: it starts where corporate teams have questions today and descends to the technical layer those questions rest on.
-
Start with AEO: the conceptual map is built here
AEO (Answer Engine Optimization) is the work of making a page directly quotable in search engine answer boxes and AI assistants (definition in the glossary). It comes first because the other three subjects only become arguable once this frame exists: without it, structured data and performance look like maintenance tasks rather than visibility decisions.
The guide covers the criteria answer engines apply when selecting a passage, how a quotable passage is written, and how the work is measured over a fixed question set.
-
Continue with GEO: being quoted and being cited are not the same thing
GEO (Generative Engine Optimization) is the work of being cited as a brand and a page inside answers produced by generative engines — ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews (definition in the glossary). The difference from AEO is the channel: what decides the outcome here is not passage quality alone but crawler access and entity clarity.
The guide sets out both sides of the AI crawler access decision without arguing for one. Our own robots.txt and llms.txt files are open as live examples.
-
Go down to Core Web Vitals: a page that cannot be loaded cannot be quoted
Everything in the first two guides depends on the page loading quickly and behaving predictably. Core Web Vitals are the three measurable user experience metrics Google publishes: LCP for loading, INP for interaction response and CLS for visual stability.
The guide gives the published thresholds for all three, explains why field data and lab data disagree, and sets the order in which fixes are applied. This site's own measurements are embedded as a worked example.
-
Close the foundation with Schema.org: this is where a machine learns what your company is
Structured data is the practice of marking up the information on a page with a standard vocabulary that search and generative engines can read without guessing; on corporate sites the JSON-LD format is used. What your company is, where it operates and what it delivers reaches the machine through this layer.
The guide holds a type-selection decision table, the logic of an
@id-based entity graph, validation tools and the eight mistakes we see most often, with copyable JSON-LD examples. -
Keep the glossary open: all four guides speak the same vocabulary
The glossary was written to be kept open while reading, not to be read end to end. Every term has its own anchor, and technical expressions in the guides link to those anchors the first time they appear. Definitions are short and self-contained — the length you could quote in a meeting.
If someone on your team has just taken this subject over, this is usually the right entry point: without the vocabulary in place, the guides read harder than they need to.
What is in the knowledge centre?
Five resources: guides on AEO, GEO, Core Web Vitals and Schema.org, plus the digital visibility glossary. The four guides are long form; each carries a definition box, question headings, implementation steps, common mistakes and a checklist you can copy. The fifth is an anchor-based index of terms.
- 01 What Is AEO? Why an answer engine selects one passage over another, how a quotable passage is written, which schema types serve the work and how results are measured.
- 02 What Is GEO? How generative engines choose a source, both sides of the AI crawler access decision, the correct size for llms.txt and how citation visibility is sampled.
- 03 Core Web Vitals Guide Definitions of LCP, INP and CLS with published thresholds, the difference between field and lab data, the order of fixes and how a performance budget is set.
- 04 Schema.org and JSON-LD Guide Structured data fundamentals, type selection on a corporate site, entity graphs built on @id, multi-location structures, validation and the eight frequent mistakes.
- 05 Digital Visibility Glossary Terms across search, performance, technical infrastructure, structured data and artificial intelligence, each with its own anchor and a definition that stands alone.
Reading times are estimates based on the length of each guide, not measured values; that is why they are given as ranges. The update dates are not estimates: each one is identical to the dateModified field in that page's structured data. Every guide also exists in Turkish at the matching URL, reachable from the language switch in the header.
Looking for the meaning of a single term?
The twelve terms below are the ones corporate teams meet most often in agency proposals and technical reports. Each link goes straight to the full definition in the glossary. Those definitions stand on their own: they can be pasted into an email or a meeting note without losing their meaning.
| Term | In one sentence |
|---|---|
| AEO | The work of making a page directly quotable in answer boxes and AI assistants. |
| GEO | The work of being cited as a brand and a page inside answers produced by generative engines. |
| Core Web Vitals | The shared name of the three measurable user experience metrics Google publishes: LCP, INP and CLS. |
| LCP | The time it takes for the largest content element in the viewport to render; loading speed as the user experiences it. |
| INP | The delay before a page responds visually to user interactions; the metric that replaced FID. |
| CLS | The total measure of unexpected content movement while a page loads; the visual stability metric. |
| JSON-LD | The recommended markup format, which declares structured data in a script block separate from the page body. |
| Entity | A unique object search and generative engines recognise — a company, person, product or place — whose identity is pinned with @id. |
| Extractable passage | A short block of text that keeps its meaning when lifted out of context, and can therefore be quoted directly. |
| llms.txt | A plain text file published at the site root that summarises the site's structure and main resources for language models. |
| hreflang | The tag that declares which language and country a page is written for, and where its other language versions live. |
| canonical | The tag that declares which address is the principal version when the same content is reachable at more than one URL. |
If the term you need is not listed, open the full glossary; terms are ordered by category and alphabetically.
What do we base the accuracy of this material on?
On four rules: every number is given with its source and access date, the update date shown on the page is identical to the field in the structured data, quoting with attribution is explicitly permitted, and a direct address for error reports is published. A number without a source is not written.
No unsourced number is published
Where we state a threshold, a ratio or a duration, its source and access date sit next to it. The sources we draw on are limited to public primary documentation: Google Search Central, web.dev, schema.org and the W3C accessibility criteria. We do not use second-hand statistics whose origin cannot be traced, which is why some sections here are shorter than the equivalent pages elsewhere.
The date is visible and real
A last updated date is visible on every page and is identical to the dateModified field in its structured data. If the content did not change, the date does not change either; we do not refresh dates to make a page look current.
Quoting is permitted
You may quote this material with attribution and a link; the conditions are written on the terms of use page. AI crawlers are granted the same permission explicitly: our robots.txt file shows it line by line.
Corrections come straight to us
If you believe a definition is wrong, a source has aged out or a threshold has moved, write to bilgi@bakibilisim.com. We make the correction and move the update date on that page. This material is a public reference rather than part of a product, so its errors are corrected in public too.
200 ms
The threshold Google publishes as “good” for Interaction to Next Paint (INP). Like every other number on this page, it is given with its source and access date.
Source: Google web.dev — Interaction to Next Paint (INP) · Accessed: 2026-07-29
The same rule governs our commercial pages. We publish our own site's measured scores together with the tool and the measurement date on the proof page; we have no published corporate case study yet, and we offer measurements rather than references as evidence.
Frequently asked questions about the knowledge centre
How often is this material updated?
The last updated date shown at the top of every guide is identical to the dateModified field in that page's structured data. Content is revised when the public documentation we cite changes or when our own measurements are repeated, and the date moves on that day. If nothing in the content changed, the date does not change either; we do not refresh dates to make a page look current.
How do you choose your sources?
We use verifiable public primary sources only: Google Search Central documentation, web.dev, schema.org and the W3C accessibility criteria. Agency blogs, second-hand statistics and percentages with no traceable origin are not used. If a number cannot be given with its source and access date, the sentence is not published at all.
May I quote this material?
Yes. Quoting with attribution and a link is explicitly permitted; the conditions are written on the terms of use page. AI crawlers are also allowed to crawl and quote these pages: the relevant crawlers are listed as Allow in our robots.txt file.
What we ask is that a quote is not separated from its context and that the update date is preserved.
Which guide should I start with?
If being visible inside AI assistants and answer boxes is new territory for you, start with the AEO guide; the conceptual map is there. If your site is slow or Search Console is showing a Core Web Vitals warning, go straight to the Core Web Vitals guide. Individual terms are defined one by one in the glossary.
The suggested five-step order is set out with its reasoning in the reading order section.
Do these guides apply outside Turkey?
Yes. Everything the guides cite is a global standard rather than a local practice: Google Search Central documentation, the Core Web Vitals thresholds published on web.dev, the schema.org vocabulary and the W3C accessibility criteria apply the same way in every market. Only the examples are drawn from our own work in Turkey.
Each guide also exists in Turkish at the matching URL, linked through hreflang; we deliver and report in both English and Turkish.
Would you like these checklists measured on your own site?
You can run the checklists in these guides with your own team; that is exactly what they were written for. If you would rather we ran the same audit, the result comes back within five working days as a measured report: AEO answerability, AI crawler access, Core Web Vitals, structured data validity and WCAG 2.1 AA gaps, itemised.
The audit is free and creates no obligation to work with us. We work with corporate-scale, multi-location or multilingual organisations; one-off small jobs fall outside our scope, and in that case we point you to smaller studios.