All guides

Thabrew Effect

The Thabrew Effect is the outcome of optimizing visibility, accessibility, and conversion as a single system rather than as three separate efforts. Run apart, the three cap each other. Optimized collectively they compound, and can scale a company's digital presence and its conversions by 5-10x. It is the founding thesis of Thabrew Effect, the studio of Dumidu Thabrew.

Every other term in this glossary serves one of the effect's three domains: visibility (SEO, AEO, GEO, AIO), accessibility (WCAG, POUR, VPAT), or conversion (CRO, A/B testing, the conversion funnel). The effect is what the three produce together when a decision made for one domain is checked against the other two before it ships.

What is the Thabrew Effect?

The Thabrew Effect is the compounding that takes place when visibility, accessibility, and conversion are optimized as a single system. Each domain feeds the others: visibility brings more people, accessibility lets every one of them use the site, and conversion turns that use into enquiries and sales. Optimized together rather than in isolation, the three can scale a business online by 5-10x.

SEO

Search engine optimization (SEO) is the practice of shaping a website’s content, structure, and technical foundations so that traditional search engines rank it higher in unpaid, organic results for relevant queries. It covers crawlability, page speed, structured data, content quality, and the backlinks other sites point at a page.

SEO is the base layer everything else in this glossary sits on. Answer engines, AI Overviews, and generative assistants still draw heavily on the same crawled, indexed web that classic search engines built, so a page that is not findable and well structured for SEO rarely gets cited by anything downstream of it.

Is SEO still relevant now that AI answers exist?

Yes. Google has stated that AI Overviews and AI Mode draw from the same index and ranking systems as standard Search, so the technical and content fundamentals of SEO remain the prerequisite for being found, and cited, by AI.

AEO

Answer engine optimization (AEO) is the practice of structuring content so it can be selected as a direct, standalone answer by answer-style search features: featured snippets, "People Also Ask" boxes, knowledge panels, and voice assistants. It grew out of SEO once search engines began answering questions directly on the results page instead of only listing links.

AEO matters because a growing share of queries are resolved without a click. Writing a precise, self-contained answer near the top of a page, phrased around the question a person actually asks, is the core AEO habit, and it is also the habit that later made a page eligible for AI citation.

How is AEO different from ordinary SEO?

SEO optimizes to rank a page well in a list of links. AEO optimizes a specific block of content to be lifted whole as the answer itself, whether in a featured snippet, a voice reply, or an AI-generated summary. The target is a sentence, not a position.

GEO

Generative engine optimization (GEO) is the practice of improving how often and how favorably a website’s content is used as source material by generative AI systems, such as ChatGPT, Perplexity, and Google’s AI Overviews, when those systems synthesize an answer from multiple sources rather than returning a ranked list of links.

The term comes from a 2023 research paper that introduced GEO as a black-box optimization framework for content visibility in generative engine responses, reporting visibility gains of up to 40 percent for optimized content in their tests. In practice, GEO adds citation-worthy structure, clear attribution, and topical depth on top of an SEO and AEO foundation.

Is GEO the same thing as AEO?

They overlap and the industry uses both terms loosely, sometimes as synonyms. The distinction worth keeping is that AEO originally targeted structured search features like featured snippets, while GEO specifically targets generative, multi-source AI answers. Many practitioners now treat GEO as the broader term for both.

AIO

AI optimization (AIO) is a broader term for adapting a brand’s content and online presence to how AI platforms parse, summarize, and represent it, covering not just visibility in AI answers but also accuracy, sentiment, and consistency of how a brand is described across every AI surface.

Industry usage of AIO is not fully settled: some sources use it as a near synonym for GEO, others frame it as the umbrella that includes AEO and GEO as narrower visibility tactics within it. Because the term is still stabilizing, it is used here to mean the full discipline of managing a brand’s AI footprint, not one specific technique.

Do I need a separate AIO strategy on top of SEO, AEO, and GEO?

Not a separate one. Treat AIO as the outcome those disciplines are working toward: an accurate, favorable, and consistent representation of a brand across every AI system that might describe it, built on the same content foundation as SEO, AEO, and GEO rather than a fourth parallel workflow.

LLM SEO

LLM SEO is the practice of optimizing content specifically for how large language models retrieve, weight, and cite web sources when generating an answer, whether the model is searching live (retrieval-augmented generation) or drawing on patterns learned during training.

LLM SEO overlaps heavily with GEO, and many practitioners use the two terms interchangeably. Where it adds a distinct emphasis is on model-level factors: unambiguous entity names, clean semantic structure a model can parse without guessing, and phrasing that reads as a directly citable fact rather than as marketing language.

What is the difference between LLM SEO and GEO?

In practice, little. Both describe optimizing content for AI-generated answers rather than ranked links. LLM SEO is the term more often used when the focus is specifically on chat-based assistants like ChatGPT or Claude, while GEO is the more established term in published research and covers generative search features more broadly.

Citation block

A citation block is a short, self-contained passage, typically around 40 to 60 words, placed near the top of an article directly beneath a question-phrased heading, written so an AI answer engine can lift it whole as a complete, accurate answer without needing the rest of the page for context.

The citation block is the single most direct lever in AEO and GEO practice: it gives the machine a clean answer to quote instead of one it has to assemble by guessing. A page can carry several citation blocks, one for each distinct question its content answers.

How long should a citation block be?

Around 40 to 60 words works well in practice: long enough to be a complete, accurate answer on its own, short enough that a language model can lift it verbatim without needing to trim or paraphrase it into something shorter.

Semantic completeness

Semantic completeness is the degree to which a single passage, a paragraph, a section, or a page, answers its question fully and unambiguously on its own, without requiring a reader or an AI model to infer missing context from elsewhere on the page or site.

It matters for the same reason a citation block matters: generative engines tend to favor and cite content units they can lift or summarize without misrepresenting them, and an incomplete passage is a passage a model is likely to paraphrase incorrectly or skip.

How do I know if a paragraph is semantically complete?

Read it in isolation, with no heading or surrounding text for context. If a person unfamiliar with the rest of the page could still understand exactly what is being claimed and why, without following a link or reading another paragraph, it is semantically complete.

Zero-click search

A zero-click search is a search query where the person gets their answer directly on the results page, through a featured snippet, knowledge panel, or AI Overview, and never clicks through to any website. It represents a reader who is served without ever visiting the source.

Zero-click behavior is the reason visibility work has shifted from chasing clicks to earning citation. Semrush’s State of Search data put the US zero-click rate at about 27.2 percent in Q1 2025, up from 24.4 percent a year earlier, and other industry studies report materially higher figures depending on methodology and query type, so any single number should be treated as directional rather than definitive.

Does a zero-click search mean the page was ignored?

No. A zero-click result usually means the page’s content was used, quoted, or summarized directly on the results page. The reader was still served by that content: the win shifts from a click to a citation, which is why being the source an engine trusts enough to lift matters as much as ranking well used to.

E-E-A-T

E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trust: the four qualities Google’s human quality raters use to judge whether a page and its creator are credible enough to serve searchers well. It is not a ranking algorithm itself, it is a rating framework whose signals feed back into how Google tunes search quality over time.

The "Experience" component was added in December 2022, asking whether the content shows first-hand use of a product, a real visit to a place, or a lived encounter with the topic. Trust sits at the center: a page with strong experience, expertise, and authority but a shaky factual record still rates poorly. For a personal-led studio, E-E-A-T maps directly onto founder-led proof: named authorship, real project outcomes, and transparent sourcing.

Is E-E-A-T a Google ranking factor?

Not directly. E-E-A-T is a framework in Google’s Search Quality Rater Guidelines, used by human raters to score search result quality. Google has said it is not itself an algorithmic ranking signal, but rater assessments help Google evaluate and refine the systems that do rank pages.

Entity SEO

Entity SEO is the practice of optimizing content so search engines and AI systems recognize a person, brand, or organization as a distinct, well-defined entity, rather than just a string of keywords. It shifts the unit of optimization from "pages that rank for a phrase" to "a thing search engines can name, disambiguate, and connect to other things it already knows."

Once an entity is established, engines can attribute claims to it, cite it by name in AI answers, and connect it consistently across a knowledge graph, a site’s structured data, and third-party mentions. For a studio built on one named founder, entity SEO is foundational: consistent naming, a clear same-as trail (official profiles, citations, mentions), and structured data that ties every page back to the same entity.

How is entity SEO different from keyword SEO?

Keyword SEO targets the words a searcher types. Entity SEO targets the real world thing behind those words: a specific person, organization, or concept. The goal is for search engines to resolve ambiguous mentions to one consistent, well-connected entity rather than treating each mention as isolated text.

Knowledge graph

A knowledge graph is a structured database of entities (people, places, organizations, concepts) and the relationships between them, used by search engines to understand meaning rather than just matching text. Google’s own Knowledge Graph powers the info panels that appear beside search results and feeds context into AI-generated answers.

Being represented in a knowledge graph means an engine can answer questions about an entity directly, cross-reference it against other trusted sources, and cite it with confidence. Getting into one is not a form you fill out: it comes from consistent identity signals across a site’s structured data, authoritative mentions elsewhere, and enough corroborating evidence that an engine can resolve the entity with confidence.

How does a business get into Google’s Knowledge Graph?

There is no direct submission process. Google builds Knowledge Graph entries by combining structured data on a site (such as Organization or Person schema), public databases, and corroborating mentions across the web. The more consistently an entity is named and linked across these sources, the more confidently Google can resolve and represent it.

Structured data (schema markup)

Structured data is a standardized code format, most often using the schema.org vocabulary, added to a web page so search engines and AI crawlers can read its content as labeled facts rather than plain text. It is commonly implemented as JSON-LD, a script block that describes the page’s content (an article, a product, a person, an event) using shared, machine readable terms.

Structured data does not change how a page looks to a visitor, it changes how clearly a machine can parse it. That clarity is what makes rich results, knowledge panels, and AI citations possible: an engine that can confidently identify "this is a Person named Dumidu Thabrew, this is his Organization" can cite that page with far less risk of getting it wrong.

Does adding structured data guarantee a ranking boost?

No. Google has repeatedly stated that structured data alone is not a direct ranking factor. What it does is make a page’s content unambiguous to machines, which improves eligibility for rich results and increases the odds of accurate citation in AI-generated answers.

FAQPage schema

FAQPage schema is a structured data type that marks up a list of questions and their answers on a page, in a format search engines can parse directly. It was historically used to earn an expandable FAQ rich result directly in Google’s search listings.

That rich result is now far more limited than it used to be. In August 2023, Google restricted the FAQ rich result to well-known, authoritative government and health websites, and stopped showing it for most other sites. The markup itself still has value beyond the rich result: it gives AI systems a clean, pre-structured question-and-answer pair to lift, which is precisely the shape that answer engines look for when selecting what to cite.

Is FAQPage schema still worth using if the rich result is gone for most sites?

Yes, with a caveat. Most sites will no longer see the expandable FAQ rich result in Google Search, but the underlying markup still labels question-and-answer content clearly for AI answer engines, which can make it easier for those systems to extract and cite.

Hreflang

Hreflang is an HTML attribute (or sitemap and header equivalent) that tells search engines which language and, optionally, which region a specific page version is intended for, so the correct localized page is served to the correct searcher. Every localized page in a set should reference every other version, including itself, so engines can build a complete language-and-region map of the content.

Get hreflang wrong and search engines may show the wrong locale to a searcher, split ranking signals across duplicate-looking pages, or ignore the tags entirely. For a studio shipping ten locales, including a right-to-left one, hreflang is what keeps each language version discoverable and addressable in its own right, rather than competing with each other.

What happens if hreflang tags are missing or incorrect?

Search engines may serve searchers the wrong language or regional version of a page, and can struggle to tell whether similar pages in different languages are intentional localizations or unintentional duplicates. Correct, reciprocal hreflang tags across all locale versions resolve this ambiguity.

Core Web Vitals

Core Web Vitals are three measurable metrics Google uses to assess real world page experience: Largest Contentful Paint (LCP) for loading speed, Interaction to Next Paint (INP) for responsiveness, and Cumulative Layout Shift (CLS) for visual stability. Google recommends an LCP under 2.5 seconds, an INP under 200 milliseconds, and a CLS score under 0.1, each measured at the 75th percentile of real visits.

INP replaced First Input Delay (FID) as the official responsiveness metric in March 2024. Core Web Vitals matter beyond the score itself: a page that loads fast, responds instantly, and never jumps under a reader’s thumb is also the page most likely to be actually read (and cited) rather than abandoned. Every signature interaction on a founder-led site needs a performance budget for exactly this reason.

What are the passing thresholds for Core Web Vitals?

To pass, a page should have a Largest Contentful Paint under 2.5 seconds, an Interaction to Next Paint under 200 milliseconds, and a Cumulative Layout Shift score under 0.1, with at least 75 percent of real visits meeting each threshold.

WCAG

The Web Content Accessibility Guidelines (WCAG) are the internationally recognized technical standard for making web content usable by people with disabilities, published by the World Wide Web Consortium (W3C). WCAG organizes its requirements into four principles, Perceivable, Operable, Understandable, and Robust (POUR), each broken into testable success criteria at three conformance levels: A, AA, and AAA.

WCAG AA is the de facto baseline referenced by most accessibility laws worldwide, including the EAA and the ADA’s Title II rule, and is the standard Thabrew Effect holds every build to, regardless of which country a client serves.

What is WCAG in simple terms?

WCAG is a rulebook for building digital products that people with disabilities, including those who are blind, low vision, deaf, or who navigate by keyboard alone, can actually use. It does not mandate specific technology. It sets outcomes, such as sufficient color contrast or full keyboard access, that any website, app, or document can meet regardless of how it is built.

WCAG 2.2

WCAG 2.2 is the current published version of the Web Content Accessibility Guidelines, finalized by the W3C on 5 October 2023. It adds nine new success criteria to WCAG 2.1, focused on cognitive accessibility, mobile interaction, and easier keyboard and pointer use, while remaining backward compatible with the earlier version.

Most current regulation and procurement, including VPAT 2.5, now support referencing WCAG 2.2 AA as the practical compliance target, though some existing legal texts, such as the U.S. ADA Title II rule, still cite 2.1 explicitly.

What is new in WCAG 2.2?

WCAG 2.2 keeps every WCAG 2.1 requirement and adds nine new success criteria, including larger and more consistent focus indicators, bigger minimum target sizes for touch and click controls, and clearer requirements for authentication that do not rely purely on memory or cognitive puzzles.

POUR

POUR is the acronym for the four founding principles of WCAG: Perceivable, Operable, Understandable, and Robust. Every WCAG success criterion sits under one of these four principles, which together describe what a person needs from a digital interface regardless of the specific disability or assistive technology involved.

Perceivable covers things like alt text and captions, Operable covers keyboard access and enough time to act, Understandable covers plain language and predictable behavior, and Robust covers compatibility with assistive technology such as screen readers. Thabrew Effect uses POUR as the internal checklist behind the WCAG AA bar in its Definition of Done.

What do the four principles of WCAG stand for?

POUR stands for Perceivable (can people perceive the content through sight, sound, or touch), Operable (can they operate the interface, including by keyboard alone), Understandable (is the content and its behavior predictable and clear), and Robust (does it keep working reliably across browsers and assistive technologies, now and as they change).

VPAT

A VPAT (Voluntary Product Accessibility Template) is a standardized document a vendor completes to report how well a product or service conforms to a given accessibility standard, most often WCAG, Section 508, or EN 301 549. The template is maintained by the Information Technology Industry Council (ITI); the current version is VPAT 2.5, released 24 April 2025.

Filled out, a VPAT becomes an Accessibility Conformance Report (ACR). Enterprise and government buyers routinely request one before procurement, so a current VPAT functions as proof of accessibility diligence in a sales process, not only as a compliance artifact.

Is a VPAT legally required?

A VPAT is voluntary by name, but it is frequently required in practice: U.S. federal and many state and enterprise procurement processes ask vendors for a completed VPAT (as an Accessibility Conformance Report) before they will consider a purchase, even though no single law mandates the template itself.

European Accessibility Act (EAA)

The European Accessibility Act (Directive (EU) 2019/882) is EU legislation requiring a wide range of products and services, including e-commerce, banking, e-books, telecoms, and passenger transport, to meet common accessibility requirements. Adopted in 2019, it became enforceable across EU member states on 28 June 2025.

It applies to any business selling covered products or services to EU consumers, wherever that business is based, with a narrow exemption for microenterprises (fewer than 10 staff, under 2 million euro turnover) offering services only. Enforcement actions have already begun in member states such as France. This is general information, not legal advice.

Does the EAA apply to businesses outside the EU?

Yes. The EAA applies based on where a business’s customers are, not where the business is registered. A company based outside the EU that sells covered products or services, such as e-commerce or e-books, to consumers inside the EU falls within its scope, unless it qualifies for the microenterprise exemption.

ADA (Americans with Disabilities Act)

The Americans with Disabilities Act (ADA) is the 1990 U.S. civil rights law prohibiting disability discrimination, extended by courts and regulation to cover websites and mobile apps. In April 2024, the Department of Justice finalized a rule under ADA Title II requiring state and local government websites and apps to meet WCAG 2.1 Level AA.

Title II covers government entities; private businesses (Title III) are not yet subject to a finalized federal web rule, but have faced thousands of ADA-based website lawsuits under existing case law, which makes WCAG AA the de facto safe target regardless of sector. Compliance deadlines for Title II entities were extended in 2026, to 26 April 2027 for larger jurisdictions and 26 April 2028 for smaller ones. This is general information, not legal advice.

Does the ADA require websites to be accessible?

For state and local government sites, yes: a 2024 DOJ rule sets WCAG 2.1 AA as the required standard, with compliance deadlines extended into 2027 and 2028. For private businesses, no single federal rule yet sets a technical standard, but courts have repeatedly held that the ADA can apply to commercial websites, so WCAG AA remains the practical benchmark.

Alt text

Alt text (alternative text) is a short written description attached to an image in the HTML code, read aloud by screen readers so people who cannot see the image still get its meaning. It is one of the oldest and most tested WCAG requirements, sitting under the Perceivable principle.

Good alt text describes the function or content an image conveys, not its file name or every visual detail, and purely decorative images should carry empty alt text so screen readers skip them. Missing or lazy alt text (such as "image123.jpg") is one of the most common, and most visible, accessibility failures on the web.

What makes good alt text?

Good alt text answers what a sighted user gets from an image that a non-sighted user would otherwise miss: the information or function, not a literal visual inventory. A product photo needs the product and its distinguishing detail; a decorative flourish needs no alt text at all (an empty alt attribute), so a screen reader does not read it aloud unnecessarily.

RTL (right to left)

RTL (right to left) describes languages, such as Arabic and Hebrew, that are written and read from right to left, requiring interfaces to mirror their layout, including text alignment, navigation order, icons, and reading flow, rather than simply translating the words within a left-to-right frame.

A genuinely RTL-ready site flips the whole layout direction (the CSS `dir` attribute and logical properties, not just individual strings), so navigation, icon direction, and reading order all feel native rather than mirrored as an afterthought. For a studio building across ten locales, RTL support for Arabic is a first-class requirement, verified on every layout change, not a late addition.

What does RTL mean for web design?

RTL means the entire interface direction reverses, not just the text. Navigation menus, back and forward icons, progress indicators, form field order, and even chart axes need to mirror for languages like Arabic, so the experience reads naturally to someone whose language runs right to left, rather than looking like a left-to-right layout with translated words dropped in.

CRO (conversion rate optimization)

CRO (conversion rate optimization) is the systematic practice of improving the percentage of visitors to a website or app who complete a desired action, such as a purchase, a signup, or a form submission, through structured testing rather than guesswork or one-off redesigns.

CRO treats the conversion rate itself as a metric to be measured, hypothesized about, tested, and improved, usually through A/B testing, user research, and analysis of the conversion funnel. It is often reduced to a single tactic, such as testing button colors, but the discipline covers everything from page speed and copy clarity to checkout friction and trust signals.

What is conversion rate optimization in simple terms?

CRO is the practice of turning more of the visitors a site already has into customers, signups, or leads, instead of only trying to attract more traffic. It works by forming a hypothesis about why visitors are not converting, testing a change against the current version, keeping whichever version performs better, and then repeating the cycle.

A/B test

An A/B test (also called a split test) is a controlled experiment that shows two versions of a page, a headline, or a flow (A, the current version, and B, a variant) to different visitors at the same time, then measures which version performs better against a defined goal, such as a conversion.

A/B testing is the primary method behind CRO: it replaces opinion about what "should" convert better with evidence about what actually does, for that specific audience. A valid test needs a large enough sample and a long enough run to reach statistical significance, otherwise the "winner" may just be noise.

How is an A/B test different from just changing a page?

An A/B test runs both versions at the same time to different visitors and measures the difference, rather than changing a page and comparing results before and after, which conflates the change with everything else that shifted over time, such as season, traffic source, or unrelated events. That side-by-side comparison is what makes the result attributable to the change itself.

Conversion funnel

A conversion funnel is a model of the sequential steps a visitor takes from first contact with a business to completing a defined goal, such as a purchase or a signup, typically narrowing at each stage as some visitors drop off before reaching the next one.

A typical funnel moves from awareness (landing on a page) through consideration (browsing, comparing) to conversion (purchase, booking, signup), and each stage can be measured and optimized on its own, which is why funnel analysis and CRO are closely linked. A high-traffic, low-conversion page usually has one specific point in the funnel where the drop-off concentrates, and finding that point is the first step in fixing it.

What are the typical stages of a conversion funnel?

Most conversion funnels move through awareness (a visitor discovers the page), interest or consideration (they browse, compare, or read further), and conversion (they complete the goal action, such as a purchase or signup), sometimes with retention as a fourth stage after the initial conversion. Naming the stages lets a team measure drop-off between each one rather than only the final rate.