Search terms, from crawling to AI answers

SEO, or search engine optimization, is the work that helps a site's pages rank in search results for the queries its customers use: technical health, site structure, content, internal links and links from other sites. Results build over months, not weeks. In 2026 SEO is also the foundation for AI search, since AI answers draw on pages that search engines can crawl and trust. Our SEO work starts from $1,500 a month, and SEO for online stores from $2,500 a month.

More: SEO & Website Promotion

Search intent is the goal behind a search query: to learn something (informational), to compare options (commercial) or to buy or act (transactional). A page ranks and converts when it serves that goal rather than repeating the keyword: informational intent wants depth, commercial intent wants comparisons and proof, transactional intent wants a short path to the order. One query can hide several intents at once, such as cost, scope and timeline, and the same words can mean different things in different markets. Optimizing for the wrong intent brings visitors who were never going to buy.

More: SEO, AEO, and GEO in 2026: What Actually Matters to Google, and What Your Marketer's Report Should Show

Keyword research is finding the words and questions customers actually type into search, with their volume and competition, and deciding which page should answer each of them. The result is a map of queries to pages rather than a list: every article or category targets one keyword or topic cluster, with the main phrase in the title, the first paragraph, at least one subheading and the meta description. Before paying for content, it is worth asking what monthly search volume each targeted term really has.

More: Why Your E-Commerce Store Needs a Blog in 2026 — And How to Start One That Actually Works

Long-tail keywords are specific, low-volume search queries, such as a product with its model, size and material, that each bring few visits but together make up much of a large site's search traffic. They tend to convert well because the searcher already knows what they want. In a big catalogue, the products with the most revenue or demand are described by hand, and templates built from real specifications cover the long tail. Mass-produced pages can rank for hundreds of such queries for a short while and then lose them.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

Competitor gap analysis is comparing the queries, topics and formats competitors rank for with your own in order to find search demand you are missing. Tools such as Ahrefs or Semrush list the keywords where a competitor appears and you do not. The useful output is not a copy of their content but an angle every competitor ignores or a format they handle poorly. A light check every month keeps the picture current; a yearly review usually comes too late.

More: Why Your E-Commerce Store Needs a Blog in 2026 — And How to Start One That Actually Works

Organic traffic is the visits a site receives from unpaid search results, as opposed to ads, social media or direct visits. It is a means, not a goal: in one case organic traffic rose 80% while qualified leads rose 2%, which meant the content was attracting the wrong people. That is why organic traffic is read together with leads and revenue from search. New pages are also checked for whether they bring new visits or only take them from pages that already ranked.

More: The Problem You Came With Is Rarely the One You Have

Branded queries are searches that include the name of a company or its product; non-branded queries describe a need without naming anyone. Brand traffic shows that people already know the business, while growth in non-branded traffic shows that it is becoming visible to people who do not. Since March 2026, Search Console offers eligible sites a branded queries filter that recognizes the brand in any language and with typos. How often people search for a brand by name has also correlated with visibility in AI answers more strongly than backlinks.

More: SEO, AEO, and GEO in 2026: What Actually Matters to Google, and What Your Marketer's Report Should Show

A zero-click search is a search answered on the results page itself, by an AI Overview, a featured snippet or a People Also Ask box, so the user never visits a site. A widely cited figure puts the share at 38% of Google searches. For a business it separates visibility from visits: a company can appear in AI answers hundreds of times a month for a high-intent query and receive almost no clicks. Such visibility still has brand value, but it has to be measured apart from traffic and leads.

More: Why Your Business Isn't Getting Leads From Its Website (And How to Fix It in 2026)

A featured snippet is a short direct answer, taken from one page, that Google shows above the regular results as a paragraph, a list or a table. Pages win it by answering a clear question in the first lines under a matching heading. It is a classic AEO target in traditional search and can be tracked in Search Console and in tools such as Ahrefs. With question-and-answer markup in place, featured snippets are cited as appearing within 2 to 4 weeks, earlier than citations in AI Overviews.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

Striking-distance pages are pages that already rank in positions 4 to 15 for valuable queries: close to the top results, but not yet among them. They are usually the fastest win in SEO, because a sharper title, more internal links and deeper coverage of the question can move them further than ten new articles would. That is why a sound SEO plan starts with them before any new content is commissioned.

More: SEO & Website Promotion

A Google core update is a broad change to Google's main ranking systems, rolled out over several weeks, after which rankings shift across many sites at once. It is not a penalty: pages lose ground when other pages answer the query better. In 2026 Google ran two core updates, in March and May, along with four spam updates; the March update was described as pushing down AI-heavy content without original perspective, experience or proof. Data should be compared no earlier than a week after a rollout ends.

More: SEO & Website Promotion

Scaled content abuse is Google's spam-policy term for producing many pages mainly to manipulate rankings rather than to help readers, whether the pages come from people, scripts or generative AI used without adding value. Google introduced it in March 2024, together with site reputation abuse and expired domain abuse. AI-written content is not banned as such; volume without value is. Online stores built on mass-generated articles have lost their search traffic as fast as they gained it.

More: When Everyone Can Publish 500 Articles, What Is Content Worth?

YMYL, short for Your Money or Your Life, is Google's term for topics where poor information can harm a person's health, finances, safety or legal position, such as medicine, finance and law. Google holds such pages to higher quality standards, so expertise, named authorship and accuracy weigh more than elsewhere. In our medical projects, texts that ranked but were not medically accurate were removed, doctors prepared the expert material, and the SEO content was built on top of it.

More: The case of Laduchi

E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness: the qualities that Google's search quality rater guidelines use to judge whether content deserves trust, with trust as the most important. It is not a single ranking factor, but it describes what Google's systems aim to reward: content written by people who have done the thing, with clear authors and sources.

More: Human-First Content Strategy: Why AI Content Is Losing the Trust War in 2026

An SEO report is a regular summary, usually monthly, of what search work achieved and what was done, written for the business owner. Activity counts, such as articles published or backlinks built, are not results: a useful report ends with leads and revenue from search. The structure we propose has seven parts: search visibility, growth, content, technical SEO, AI search, business results and, last, the list of work done.

More: SEO, AEO, and GEO in 2026: What Actually Matters to Google, and What Your Marketer's Report Should Show

Technical SEO is the part of SEO that makes a site readable for search engines: crawling, indexing, redirects, duplicates, canonical tags, structured data and speed for real visitors. Its job is to remove obstacles so that content and links can do their work. Site structure, page speed and internal linking are cheapest to get right when they are planned from the first day of a project rather than added later. Technical SEO has not lost its importance; it has stopped being the whole strategy.

More: SEO & Website Promotion

An SEO audit is a structured review of a site's technical state, structure, content and links that finds what holds its search visibility back and ranks the fixes by impact. A useful audit works from data rather than a generic checklist: a full crawl, Search Console, analytics and, where the hosting keeps them, server logs. Our store structure audit for online stores starts from $2,500: it checks categories, filters, product cards, structured data and speed, adds Merchant Center diagnostics, and ends in a plan ordered by impact.

More: SEO & Website Promotion

A search crawler, or bot, is a program that search engines and AI services send to fetch and read web pages, such as Googlebot, Bingbot or OpenAI's OAI-SearchBot. Googlebot renders pages in an up-to-date headless Chromium and crawls mostly from the US without a preferred language. If a firewall or a hosting rule turns a crawler away, the content effectively does not exist for that search engine or assistant, even though people see it as usual. Third-party crawlers are also used in audits, for example to check hreflang across a whole site.

More: The Problem You Came With Is Rarely the One You Have

Crawl budget is the number of URLs on a site that a search engine can and wants to crawl in a given time; Google defines it through crawl capacity and crawl demand. Google says it matters mainly for sites with about a million pages, or with ten thousand pages that change daily. In large online stores, filter combinations and duplicate URLs are what usually waste it. Blocking unimportant URLs in robots.txt and returning 404 or 410 for removed pages helps; noindex does not, because Google still has to request a page to see the tag.

More: E-Commerce SEO

Indexing is the step in which a search engine stores a crawled page in its index, the database from which search results are drawn; a page that is not indexed cannot rank for anything. Crawling does not guarantee indexing: duplicates, thin pages and pages marked noindex are left out. Unwatched crawl errors and sitemap problems let pages slip out of the index even on a site that publishes a lot. In one of our audits, 350 pages entered the index after the fixes.

More: The Problem You Came With Is Rarely the One You Have

robots.txt is a text file at the root of a site that tells crawlers which addresses they should not fetch. It stops crawling, not indexing: a blocked page can stay in the index, and any noindex or canonical tag on it becomes invisible, because the crawler never reads the page. The rules are advisory, and not every bot follows them. Hosting providers sometimes add their own disallow rules unnoticed, so the file is worth checking after every move or upgrade.

More: E-Commerce SEO

Noindex is a directive, in a robots meta tag or an HTTP header, that keeps a page out of the search index while still letting search engines crawl it and follow its links. It works only if the page is not blocked in robots.txt, and it does not save crawl budget. In online stores, sort orders, internal search results and an empty category are typical candidates. On a live site a single forgotten noindex can remove it from Google, which is why we close test copies at the server level instead.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

A canonical URL is the address that a site declares, with a rel="canonical" tag, as the main version of a page available under several URLs, so that search engines consolidate signals there. Google treats it as a strong hint, not a command. It does not stop crawling and is not interchangeable with noindex or robots.txt; mixing them on the same pages is a common SEO mistake.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

Duplicate content is the same or nearly the same content available under several addresses, on one site or across many. Search engines then have to pick one version, and ranking signals spread across the copies, so none of them ranks as well as a single page would. Typical sources in online stores are filter and sort combinations, conflicting SEO modules installed by different developers, and manufacturer descriptions copied by dozens of shops. Canonical tags, clean URL rules and original text are the usual remedies.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

Keyword cannibalization is a situation in which several pages of one site compete for the same search query, so search engines alternate between them and none ranks as strongly as one clear page would. In online stores it often comes from near-identical filter pages competing with each other and with their own category; on blogs, from several articles on the same question. The fix is to decide which page owns the query, then merge, re-target or canonicalize the others.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

Thin content is pages with little real value for the reader: near-empty templates, auto-generated pages or texts that repeat what already exists elsewhere. Besides not ranking themselves, many such pages weigh on how search engines judge the whole site. A language version that only says the article exists in another language is a typical case, and so are machine-translated versions nobody reviewed. Ten thousand auto-generated thin pages are technical debt dressed up as an asset.

More: The Problem You Came With Is Rarely the One You Have

Index bloat is a search index swollen with low-value addresses of one site, most often filter, sort and pagination combinations, each of which creates a new URL. One shoe category with ten colours, twelve sizes, thirty brands and five price ranges can produce about 18,000 addresses, and crawlers spend their time on these near-duplicates while important pages wait. The cure is a rule written into the code before launch: which combinations get a page and which never become an indexable URL. An AI model given a vague brief once turned every product parameter into its own address.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

SEO filter landing pages are filter combinations, such as one brand within one product type, that a store deliberately opens to indexing because people search for exactly that combination. From the search side the decision rests on five checks: its own query, enough products, a stable assortment, an intent different from its category, and a results page where such pages actually rank; in one project a combination is indexed only with at least six products. Approved pages get a self-referencing canonical and a sitemap entry, while the other combinations are disallowed in robots.txt and return 404 when empty, as Google's guidance updated in December 2025 advises. In one case, selected brand and product type pages grew from 263 to 670 organic clicks a month.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

URL structure is the pattern of a site's addresses: folders, language prefixes, how categories, products and articles are named, and which parameters exist. Readable, stable addresses help both visitors and search engines see where a page sits. Because URL rules, locale prefixes and indexability shape the page templates themselves, we fix them before navigation and templates are designed. Changing the structure after launch means a full migration with redirects.

More: The Website Design Process Most Agencies Show You Is Wrong

URL normalization is making sure that each page is reachable under exactly one address. Without it, HTTP and HTTPS, www and non-www, a trailing slash, capital letters and parameters in a different order each count as separate URLs with separate signals. In online stores, filter parameters should always appear in the same order, so that one selection has one address, and sort variations, which show the same products, should never be indexed. These rules are tested before launch together with pagination, canonical and robots settings.

More: Filters and Faceted Navigation: How Online Stores Create Thousands of Duplicate Pages

JavaScript SEO is making sure that content built or changed by JavaScript can still be rendered, read and indexed by search engines. Googlebot renders pages in a headless Chromium and reads the resulting HTML, but rendering takes time and can fail, so headings, text, links and prices are safest in the HTML the server sends. On our own 3D home page all of them are plain HTML delivered before any script runs. Filters that change products without changing the address create no duplicates, but they also create no page that can rank.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

Cloaking is showing search engines different content from what people see, usually by detecting the crawler's user agent or IP address. It violates Google's spam policies and can lead to a site's removal from results. A lighter version of a page chosen in the browser according to what the device can handle, with the same HTML sent to everyone, is not cloaking. Our own home page works that way, giving devices without a graphics card a simpler first screen.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

An XML sitemap is a file that lists the addresses a site wants search engines to crawl, usually with the date of their last change. It does not force indexing, but it helps crawlers find new and updated pages, especially deep in a large catalogue. It should list live canonical pages only, so approved filter landing pages belong there, while redirects and noindex pages do not. Keeping it current and submitting it in Bing Webmaster Tools as well as Google Search Console also matters for AI search.

More: E-Commerce SEO

A 404 status tells browsers and search engines that a page does not exist, 410 says it is gone for good, and a soft 404 is an empty or missing page that wrongly answers with status 200, as if everything were fine. Removed pages and filter combinations with no products should return 404 or 410, not an empty page and not an automatic redirect to the main category. Soft 404s waste crawling and confuse indexing. After every migration, new 404 errors are watched closely, since they show which redirects were missed.

More: E-Commerce SEO

Server log analysis is reading the hosting's access logs to see which pages search bots actually request, how often and with which response codes. Crawling tools show what could be crawled; logs show what Googlebot and other bots really spend their time on. We use it in our store structure audit wherever the hosting keeps logs, and together with Search Console and analytics it shows which filter landing pages to add, keep or remove.

More: E-Commerce SEO

IndexNow is an open protocol that lets a site notify search engines as soon as a page is added, changed or deleted, instead of waiting to be crawled. One submission is shared among the participating engines: Bing, Yandex, Naver, Seznam.cz, Yep and Amazon (IndexNow, October 2026). Google is not among them, so for Google sitemaps and its own crawling still do the work. We submit our own new pages through IndexNow every night.

More: SEO & Website Promotion

Core Web Vitals are three Google metrics of real user experience: loading (Largest Contentful Paint, LCP), responsiveness (Interaction to Next Paint, INP) and visual stability (Cumulative Layout Shift, CLS). Google judges them on field data from real visits at the 75th percentile, and the good thresholds are 2.5 seconds or less for LCP, 200 milliseconds or less for INP and 0.1 or less for CLS. They are one ranking signal among many and a direct measure of whether visitors wait and stay. Our monthly SEO report tracks them in its technical part.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

Largest Contentful Paint (LCP) is the Core Web Vitals metric for loading: the time until the largest visible element of the first screen, usually a main image or a block of text, is drawn. Google counts 2.5 seconds or less at the 75th percentile of visits as good. A slow server, heavy images and fonts or scripts that block rendering are the usual causes of a poor LCP. Visitors on mid-range phones and mobile connections feel it most.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

Interaction to Next Paint (INP) is the Core Web Vitals metric for responsiveness: how long a page takes to react visibly to a tap, click or key press, measured over the whole visit. Good is 200 milliseconds or less. Visitors on weak devices suffer most, because heavy JavaScript keeps the browser's main thread busy. The related lab metric, Total Blocking Time, was 16 seconds on our 3D home page in software rendering and fell to 0 once such devices received a lighter first screen.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

Cumulative Layout Shift (CLS) is the Core Web Vitals metric for visual stability: how much content jumps while a page loads, for example when an image, a banner or a late font pushes the text down. Good is 0.1 or less. Shifts make people tap the wrong thing and lose their place in the text. When we moved some blocks to be built on the server rather than in the browser, CLS fell from 0.19 to 0 on our article pages and from 0.5 to 0 on the blog.

More: 3D on a Live Site: Our three.js Experiment and What Google Saw

A website migration is any change that moves a site's addresses, platform, domain, structure or language versions, carrying across URLs, content, metadata, images, internal links, structured data and redirects. What really moves is the site's accumulated search history. Rankings are most often lost through broken URLs rather than the platform change itself: one cited case lost 42% of organic traffic in 12 days without a redirect strategy. Research on 892 domain migrations tied recovery time to the size of the backlink footprint, and a redesign keeps rankings more easily than a new site.

More: The Website Migration Checklist Most Agencies Skip

A redirect map is a table that pairs every old URL of a site with its new address before a migration or a redesign, so that each one gets a permanent 301 redirect to the closest matching page. Without it, links, bookmarks and search rankings lead to 404 errors. Google advises keeping redirects for as long as possible, generally at least a year, and mass redirects to the home page can be treated as soft 404s.

More: The Website Migration Checklist Most Agencies Skip

A 301 redirect is a permanent server-side redirect that sends visitors and search engines from an old address to a new one and passes the old page's signals along. It is the basic tool of every migration and URL change. Redirects should generally stay in place for at least a year, lead to the closest matching page and avoid chains of several hops. Mass redirects to an unrelated page are a classic failure: sending 2,000 old URLs to the home page throws away what they had earned.

More: The Website Migration Checklist Most Agencies Skip

A pre-migration baseline is the record of how a site performs in search before it moves, captured while the old version is still live. It covers an inventory of every meaningful URL and its variants, analytics and Search Console data, rankings and landing pages, structured data and hreflang, and filter landing pages or old categories with links from outside. Without it, nobody can tell after launch what was lost or whether a dip is normal.

More: The Website Migration Checklist Most Agencies Skip

Post-migration monitoring is the close watch on a site after a move: organic traffic, rankings, index coverage, crawl behaviour, 404 errors and server responses. It should last at least six weeks, longer for large or multilingual sites, because rankings often hold for several weeks and then shift, by which time the team has moved on. A normal migration dips and recovers, so the trajectory matters more than a snapshot taken three days after launch.

More: The Website Migration Checklist Most Agencies Skip

Search equity is the value a site has accumulated in search over the years: domain authority, rankings, backlinks and branded searches. It does not appear in financial statements, yet it is part of how the business acquires customers, and a careless redesign or acquisition can wipe it out. A 15-year-old domain with real backlinks can be worth more than a beautiful new site, which is why a domain change is treated as a project of its own.

More: What Happens to the Website After an Acquisition?

Domain warm-up is preparing an existing domain for a relaunch so that the new site does not start from a cold position: indexable pages filled with real content, structure fixed and the SEO work planned around the launch. For a luxury jeweller whose site had not been touched since 2015, monthly visitors went from about 500 to 2,800 in the first month and 15,000 at six months, and indexed pages grew from 60, mostly empty, to 3,500 populated product pages. Launch day is where the SEO work starts, not where it ends.

More: Why Launching a New Site Is Good for SEO: The Nicole Case

Out-of-stock and discontinued product pages raise a search question: which address and status code a product page keeps once the product cannot be bought. A temporarily unavailable product keeps its URL, status 200 and its place in the index, with OutOfStock or BackOrder in the markup, so a short gap does not destroy a page that ranks. A discontinued product with a successor gets a 301 redirect to it; with nothing comparable, the page returns 404 or 410 rather than redirecting to the category, which search engines tend to treat as a soft 404. A category left empty is set to noindex until products return.

More: E-Commerce SEO

Shopify URL structure is the fixed address pattern of Shopify stores: products always sit under /products/ and collections under /collections/, and these prefixes cannot be changed or redirected away from. Collection URLs with tag filters cannot be redirected, and a store can hold up to 100,000 redirects, or 20 million on Shopify Plus. Moving a store to Shopify therefore changes almost every address and needs a complete redirect map, while the rigid structure makes leaving the platform later harder.

More: Shopify Development

A title tag is the HTML element that names a page; search engines usually show it as the clickable headline of a result, and browsers show it in the tab. It is one of the clearest on-page signals of what a page is about. A good title runs to roughly 50 to 60 characters, puts the main subject first (for a product, its name and key attribute) and makes a specific promise the page keeps. Vague or clickbait titles may win the click but lose the visitor.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

A meta description is a short summary of a page in its HTML that search engines often show under the title in the results. It is not a direct ranking factor, but it decides whether people click. It works best at up to about 155 characters, written as a short advertisement for the page that says what the reader will get, not as a list of keywords. Search engines may replace it with text from the page when it does not fit the query.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

Headings (H1, H2, H3) are the HTML levels that structure a page: one H1 names the page, H2s mark its main sections and H3s their subsections. Search engines and AI systems use them to understand what a page covers and to pull out answers to specific questions, so question-style H2 and H3 headings are part of our AEO checklist. In our review of nine luxury home pages, five had no visible H1 and one had an empty one.

More: We Measured Nine Luxury Brand Websites. Here Is What Their Design Gets Right

Image alt text is a short description of a picture in its HTML alt attribute, read aloud by screen readers and used by search engines to understand what the image shows. It should describe the image itself rather than repeat keywords. Pictures often carry information that never appears in the copy, such as style, material or colour, so tying them to what they show helps both image search and multimodal search. The clothing store in one of our cases had no alt text at all before the rewrite.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

Internal linking is the system of links between pages of one site: menus, breadcrumbs, related blocks and links within the text. It shows search engines how pages relate and which ones matter most, and it passes authority from strong pages to the rest. A strong structure leads from topic to article, service, case study and conversion, with key service pages linked from the home page, cases, expert profiles and related services. After our own Laravel upgrade, 987 internal links were checked and none was broken.

More: SEO, AEO, and GEO in 2026: What Actually Matters to Google, and What Your Marketer's Report Should Show

An orphan page is a page that no other page of the site links to, so visitors and crawlers can reach it only through a sitemap, an ad or a direct address. Such pages are crawled rarely and receive no authority from the rest of the site. Multilingual sites are prone to them: a translated page without links in its own language behaves like an orphan even when its English original sits in a well-connected cluster.

More: Multilingual AEO: Why Translation Isn't Enough for AI Search

Breadcrumbs are a navigation trail, such as Home > Category > Product, that shows where a page sits in the site's hierarchy and links back to each level. They help visitors move up through a catalogue and help search engines understand the structure; with markup they can also appear in search results. One clothing store removed them for a cleaner look, session recordings showed that customers had been using them, and they came back.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

A topic cluster is a group of linked pages around one subject: a pillar page that covers the topic broadly and shorter pages that each answer one specific question, all linked to the pillar and to each other. The structure shows search engines the depth of coverage and gives each question its own page to rank. For AEO it is recommended over a single very long page that tries to cover everything, since AI systems look for answers to specific questions.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

Topical authority is the degree to which search engines and AI systems treat a site as a reliable source on a subject, based on a connected body of content covering related questions, subtopics and context. It is built by sustained coverage, not by one big article. A translated page does not inherit the authority of its original, so each language builds its own. It also fades: every year without new content on a topic reduces it.

More: SEO, AEO, and GEO in 2026: What Actually Matters to Google, and What Your Marketer's Report Should Show

SEO copywriting is writing pages that rank in search and move the reader toward a decision at the same time, treating visibility, expertise and persuasion as one job. Length is not a ranking factor: a focused 800 words can beat a padded 3,000. Google no longer rewards keyword density, and generic keyword-stuffed text works against a page. The final measure of the copy is conversion, not traffic.

More: SEO Copywriting in 2026: What Actually Ranks — and What Converts

A category SEO text is the descriptive text on an online store's category page that helps both search engines and buyers understand the range. The old practice of a keyword block at the bottom of the category should give way to text that helps people choose: how to choose, how the options differ, which materials suit which use. Translated category texts are a common trap, since they keep the original market's keywords and target nothing in the new one.

More: E-Commerce SEO in 2026: What Has to Be Built Into the Store, Not Added Later

Unique product descriptions are product texts a store writes itself instead of copying the manufacturer's. A copied description says what dozens of other stores say, so it gives search engines no reason to prefer the page, while original text is one thing competitors cannot copy overnight. The practical approach is to write the top products by hand, fill templates for the rest from real specifications, and use AI drafts only from real product data checked by someone who knows the product.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

User-generated content (UGC) is content created by customers rather than the business: reviews, questions and answers, photos and forum posts. It adds text a store could never write itself, in buyers' own words and search phrases, and it keeps pages changing. On one professional platform we built, a forum added about ten thousand indexed pages in four languages. Traffic, though, follows from the value of the discussions, not the other way round.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

Content freshness is how current a page's information is, and how clearly the page shows it. Revisiting high-performing articles with new data is one of the highest-return SEO tasks, often more effective than writing new ones. For AI search it matters even more: pages are best updated every 6 to 12 months, with visible dates and named authors.

More: Why Your E-Commerce Store Needs a Blog in 2026 — And How to Start One That Actually Works

Original content is content based on first-hand experience, original research or data that exists nowhere else, as opposed to a rewrite of what already ranks. A simple test: would this page be worth creating if Google did not exist? A real case with the problem, the constraints and the result often outperforms ten generic articles, and named decisions build trust where generic claims do not. When the doctors of one clinic wrote about real cases, organic traffic grew and the appointment calendar filled.

More: When Everyone Can Publish 500 Articles, What Is Content Worth?

AI-generated content is text, images or other material produced by generative AI models. Search engines do not reject it for being machine-written, but volume without value does not convert: in one cited case, scaling from 4 to 40 AI articles a month raised organic traffic by 300% yet cut marketing-qualified leads by 15%, while 12 targeted, human-briefed articles grew them by 45%. Our own sequence puts AI in the middle: experience, research and data first, then AI drafting, expert review, editing, evidence and publication.

More: Why Most Businesses Are Using AI Wrong in Marketing — And It's Killing Their Conversions

A buying guide is an article that helps a customer choose, such as "how to choose" or "A vs B", written for the research stage before a purchase. Such posts reach buyers before they are ready to order and are among the most valuable content of a store blog, since they bring people close to a purchase and lower acquisition cost over time. They work best when they link straight to the categories and products they discuss.

More: Why Your E-Commerce Store Needs a Blog in 2026 — And How to Start One That Actually Works

A content audit is a review of every page in an existing content library: what it covers, what traffic and leads it brings, and whether it should be kept, updated or merged. It often shows that the problem is structure, not a shortage of articles. On one news platform, about two thirds of the library had little or no organic visibility, and the fix was reorganizing it rather than publishing more.

More: What If You're Solving the Wrong Problem?

Content debt is the gap that builds up when a business's expertise stays unpublished and customer questions go unanswered on its site while competitors keep publishing. Like technical debt, it grows quietly and costs more to close the longer it waits; one useful page a week adds up to hundreds of entry points in two years. In redesigns the term also covers inconsistent pages left by different agencies, where the migration has to decide which version is true.

More: The Hidden Cost of Standing Still: Why Modern Businesses Lose Competitive Advantage Long Before They Notice

Link building is the work of earning links to a site from other relevant sites, such as industry media, partners, directories and research that others cite, because search engines treat such links as signals of authority. Quality and relevance matter far more than the number of links. Buying links that pass ranking value violates Google's spam policies and can lead to lower rankings or removal from results.

More: Link Building in 2026: Why It Still Matters and How to Do It Right

A backlink is a link from another site to yours, which search engines read as a recommendation. Quality and relevance matter far more than quantity: one link from a respected publication outweighs dozens from weak sites. A site's backlink profile is a major reason to keep URLs stable or redirect them in a migration, and a translated page does not carry its original's links. For AI answers, an Ahrefs study found backlinks correlated with AI Overviews visibility at 0.218, far below brand mentions at 0.664.

More: Link Building in 2026: Why It Still Matters and How to Do It Right

Domain authority is the trust a domain accumulates over years through its links, history and content; third-party scores such as Domain Rating estimate it, but they are not Google's own metric. A strong domain helps its new pages rank faster. The focus has shifted from chasing a score to topical relevance. Site structure matters as well: subdirectories keep authority on one domain, while subdomains split it.

More: What Happens to the Website After an Acquisition?

Digital PR is creating material worth citing, such as original research, surveys, market reports, case studies and expert commentary, and putting it in front of journalists and editors so that links and mentions come as a byproduct. Unlike bought placements, it earns coverage on merit. A single feature in a respected publication can do more than dozens of weak backlinks, and the same mentions feed visibility in AI answers.

More: Link Building in 2026: What Works and What Doesn’t

Brand mentions are references to a business on other sites, with or without a link: reviews, partner articles, interviews, case studies, local news, Reddit threads and LinkedIn posts. Search engines and AI systems use them to judge how well known and trusted a brand is. In an Ahrefs study of 75,000 brands, mentions correlated with visibility in Google AI Overviews at 0.664, more strongly than backlinks, though a correlation is not proof of cause. Bought reviews and paid placements dressed up as editorial are not earned mentions.

More: SEO, AEO and GEO Are Amplifiers, Not Engines

A linkable asset is content created so that other sites want to cite it: an industry report, a statistics page, a calculator, a template, an interactive tool, original research or a comprehensive guide. Unlike a promotional page, it is useful on its own, so it keeps earning links for years after publication. It is the basis of most outreach and digital PR, because editors link to sources, not to offers.

More: Link Building in 2026: What Works and What Doesn’t

Link outreach is contacting site owners, editors and journalists to earn links or mentions for content that is relevant to their readers. Generic mass emails have given way to personal messages. Common forms are resource page outreach, offering content to curated collections of tools and guides, and broken link building: finding outdated or broken references on other sites and offering your page as the replacement.

More: Link Building in 2026: What Works and What Doesn’t

HARO, originally Help A Reporter Out, is the common name for platforms where journalists post requests for expert sources and experts answer in the hope of being quoted with a link. Winning answers respond to the exact question within hours, carry a real name and title, and offer a specific number or a contrarian detail. It works as a habit rather than a campaign: 15 minutes every morning beats occasional bursts.

More: Link Building in 2026: Why It Still Matters and How to Do It Right

Link spam is links created mainly to manipulate rankings, most often bought links that pass ranking value, which violates Google's spam policies. Private blog networks (PBN), link exchanges, thin guest posts and low-quality directories either no longer work or put the site at risk. A paid publication can still be worth it for its readers, provided the link is marked as sponsored. Lasting links come from partners, case studies and interviews.

More: SEO & Website Promotion

Local SEO is the work that makes a business visible for searches with a geographic element, such as a service plus a city or "near me", both in the map results (the local pack) and in the regular listings. Google ranks local results on relevance, distance and prominence, and there is no way to pay for a better local position. The map results sit above organic listings and in competitive markets often convert better than paid ads. Local proof, such as local cases, publications and directory listings, is stronger than a single local service page.

More: How to Rank Your Google Business Profile Higher in 2026 (Without Paying for Ads)

Google Business Profile is Google's free business listing that appears in Maps and local search results with the address, hours, reviews, photos and services. For local visibility it is the main asset: verify it, fill every field and choose the most specific primary category, described as the most important profile factor (Web Design Agency outperforms Marketing Agency). Weekly photos and posts, answered reviews, questions and answers and a full list of services keep it active. Entering a market in another language often means a separate profile with its own reviews.

More: How to Rank Your Google Business Profile Higher in 2026 (Without Paying for Ads)

NAP consistency means that a business's name, address and phone number appear identically everywhere: on the website, social profiles, directories and in LocalBusiness markup. Search engines cross-check these details to confirm that the listings belong to one real business, and even small differences such as "St." versus "Street" can dilute local signals. After a rebrand or a move, existing profiles should be updated rather than new ones created, which would leave duplicates.

More: How to Rank Your Google Business Profile Higher in 2026 (Without Paying for Ads)

Reviews as a local ranking signal means that the number, freshness and variety of a business's reviews, its overall rating and whether it answers them all affect its position in local search. Reviews are counted among the top three local ranking factors, and Google tracks whether a business responds. The simplest working system is to ask every satisfied client right after the project, with a direct link to the review form.

More: Why Responding to Online Reviews Matters in 2026 — And How to Do It Right

Location pages are pages built for a specific city or area, such as a service in a named city, to rank for local queries. A page tied to a specific market converts better than generic service text, but only if it carries real local substance: local cases, conditions and proof. Thin pages with nothing but an address, opening hours and a photo dilute a site's authority; one design studio that dropped its city pages and rebuilt around narrow specializations raised conversion by 15%.

More: Design You Don't Notice: Why the Best Design Disappears

Voice search is searching by speaking a question to a phone, a smart speaker or an assistant, which reads out a single answer or acts on it. Spoken queries are longer and phrased as questions, and together with AI Overviews, voice answers have reduced clicks to organic results. For local service businesses, FAQ content written the way people actually ask, combined with a complete Google Business Profile, has raised the number of calls.

More: How to Rank Your Google Business Profile Higher in 2026 (Without Paying for Ads)

Multilingual SEO is optimizing a site in several languages so that each version ranks in its own market. It treats the site as one business entering several search markets, not one site translated: terminology, intent, keywords, internal links, localized URLs, metadata, structured data and hreflang all have to work together in each language. By our estimates, each added market adds 20 to 30% to the SEO work and 10 to 30% to design and development.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

Hreflang is an annotation in a page's code, sitemap or HTTP header that tells search engines which language and regional versions of a page exist, so that each searcher is shown the right one. Every version must list all the others and itself, and x-default marks the fallback page. Each language version should also be its own canonical; pointing them all to English hides the translations.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

x-default is a special hreflang value that names the page for users whose language and region match none of the listed versions, typically a language selector or the main international version. Only one address in the set should carry it, the fallback destination, not every language version. It is optional, but without it search engines have to guess which version to show such users.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

International site structure is the choice of where language and country versions live in the addresses: separate country domains, subdomains or subfolders. Subfolders such as /ru/ and /ua/, which our own site uses, consolidate authority so that a link earned in any market helps all of them, and they are the simplest to maintain. Country domains give the strongest local signal but split authority, subdomains often combine the drawbacks of both, and language in URL parameters is not recommended. Whatever the choice, each locale needs its own prefix, a self-referencing canonical and hreflang to the others.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

A country-code top-level domain (ccTLD) is a domain ending tied to one country, such as .fr, .de or .com.ua. It is the clearest signal of the country a site targets: helpful at home and costly abroad, since a French domain tells search engines France, not Belgium. A neutral .com works across several countries. Search Console's country targeting setting, removed in September 2022, never applied to ccTLDs, because the domain already defines the country.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

An automatic redirect by IP or language sends visitors to a language or country version based on their location or browser settings. Google advises against redirects based on an assumed language: Googlebot crawls mostly from the US without a preferred language, so it may never see the other versions, and travellers or expats land on the wrong one. The safer pattern is a visible language switcher plus hreflang.

More: Multilingual Website Design: What Changes Beyond Translation

A language version, a regional version and a market are three levels of taking a site abroad. A language version is the site in another language; a regional version adapts it to a country, for example Arabic for the UAE; a market has its own prices, taxes, delivery and expectations, and often its own pages and categories. Confusing them produces sites that are translated but do not sell, because the offer itself was never adapted.

More: Multilingual Website Design: What Changes Beyond Translation

Localized keyword research is keyword research done separately for each market instead of translating the original list of keywords. Markets use different words, levels of specificity and intents for the same thing, and buyers reach the same purchase through different questions: about the problem, a brand, a comparison, regulation or price. A translated keyword can have no search demand at all, while the phrase local buyers actually use goes untargeted.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

Localization is adapting a product, site or content for a specific market: translation converts the language, localization converts the meaning, including formats, units, examples, imagery, trust signals and the order of information. At scale it is a pipeline rather than a task: 50,000 products in 8 markets are 400,000 states to keep correct. Every language needs an owner who creates, reviews, publishes and updates it; otherwise the English blog ends up with 50 articles, the second language with 32 and the third with 11.

More: Multilingual SEO: Why Translated Pages Do Not Rank in Their Own Markets

Transcreation is adapting content for another language and market instead of translating it word for word: the message, examples, tone, prices, units and calls to action are rewritten for local readers, while the facts stay the same. It matters for multilingual websites, ads and brand copy, where a literal translation reads as foreign and rarely ranks or sells in the target market.

More: Why Multi-Language Websites Need a Different Design Approach

Multilingual AEO is making a business understandable and citable for AI assistants in every language it serves, not only in English. Each language version needs its own links, outside sources and entity names; translation alone is not enough. A practical order of work is entity-defining pages first, then core service pages, commercial questions, proof and cases, with supporting articles last.

More: AEO & GEO Optimization

AEO, or answer engine optimization, is the work of making a business easy for answer engines such as Google AI Overviews, ChatGPT and Perplexity to understand, quote and recommend. In practice it means pages that answer real questions directly, clear facts about who the business is and what it does, structured data and consistent information across the web. AEO builds on SEO rather than replacing it. Google's own July 2026 guide puts it plainly: for Google Search, this is SEO done properly.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

GEO, or generative engine optimization, is the work of improving how generative AI systems describe and recommend a business in their answers. The term comes from a 2023 research paper; in practice GEO overlaps with AEO and covers a clear business entity, structured data, consistent language versions and content that AI can cite. Nobody can guarantee a mention in an AI answer. Our AEO and GEO audit costs $1,500.

More: AEO & GEO Optimization

An answer engine is a system that responds to a question with a composed answer instead of a list of links, such as ChatGPT, Perplexity, Gemini or Google AI Overviews. Asked for a provider or a product, it names a short list of businesses it understands well enough to trust. ChatGPT alone is cited as handling over 2 billion queries a day. For a business this means competing to be one of the few names in the answer rather than one of ten links.

More: AEO & GEO Optimization

Google AI Overviews are AI-generated summaries shown above the regular search results with links to their sources; AI Mode is Google's separate conversational search mode. AI Overviews are cited as appearing in roughly 55 to 60% of searches, and 55% of their citations come from the first 30% of a page's content, so the answer should come early. Since 2026, Search Console reports impressions from both.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

Answer-first content is writing in which the direct answer to the question of a page or section comes first, within the first 40 to 60 words, and the detail follows. AI systems extract passages, not whole pages, so each section has to answer its heading on its own: cover every section but one and check that the remaining one still makes sense. The first 200 words of a page should answer its main question completely.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

An AEO and GEO audit is a review of how AI assistants such as ChatGPT, Gemini, Perplexity and Google AI Overviews describe and recommend a business, and why. Buyer questions are written natively in each language and put to each assistant on a fixed schedule, since answers change between sessions. The results are sorted into not mentioned, mentioned but not cited, cited with the wrong page, and recommended, each with its own fix. Ours costs $1,500 and takes two to three weeks; ongoing work starts from $2,000 a month.

More: AEO & GEO Optimization

AI share of voice is the share of AI answers and citations that mention a brand, compared with its competitors, for the questions buyers ask. It depends on how densely the brand appears across the sources AI systems trust: publications, LinkedIn, Reddit and reviews. Tracked monthly as citation share across AI platforms, it turns PR, guest articles and podcast appearances into AEO tactics.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

An entity is a clearly identified thing, such as a company, person, product or place, that search engines and AI systems recognize and connect with facts: what it does, where, with whom and with what evidence. A business becomes a reliable entity when its name, description and identifiers are consistent everywhere: one canonical name with accepted variants per language, one organization identifier on every page, and updates whenever prices, people or addresses change. Transliteration multiplies names: our founder appears as Yevhen Borovoi, Eugene Borovoy, Євген Боровий and Евгений Боровой.

More: Multilingual AEO: Why Translation Isn't Enough for AI Search

llms.txt is a proposed Markdown file at the root of a website that gives large language models (LLMs) a short, curated map of its key pages. It was proposed in 2024 and is not an official standard. Google's Search documentation (checked October 2026) says that Google Search does not use these files, so they neither help nor hurt rankings; other AI services may read them. Our maintenance plans keep it current along with structured data.

More: Website Maintenance & Support

AI crawlers are the bots AI companies use to collect and fetch web content, such as GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot. Some gather training data, others fetch pages to answer a user, and blocking them has consequences: OpenAI says sites that block OAI-SearchBot are not shown in ChatGPT search answers. A firewall or hosting rule can shut them out unnoticed; one of our upgrade audits found hosting that blocked AI crawlers, and in our test of 13 US online stores, three blocked one AI crawler each.

More: AEO & GEO Optimization

Bing Webmaster Tools is Microsoft's free console for site owners, the Bing counterpart of Google Search Console: it shows indexing, queries and errors and accepts sitemaps. It matters beyond Bing itself, because ChatGPT's web search runs on Bing. Verifying a site and submitting its sitemap takes about 20 minutes and puts the site into the pool ChatGPT can cite.

More: Your Business Is Invisible to AI Search in 2026 — Here's How to Fix It

Multimodal search is search that starts from an image, the camera or a mix of image and text rather than a typed query, as in Google Lens, Circle to Search or a photo uploaded to an AI assistant. On 24 September 2026 Google added a multimodal search type to Search Console reports, without a query dimension, since there is no text query to show. To be found this way, images must be clearly tied to what they show through alt text, captions, surrounding text and markup.

More: The Search Query Is Disappearing

Structured data is code on a page, usually JSON-LD (JavaScript Object Notation for Linked Data) with the schema.org vocabulary, that tells search engines and AI systems exactly what the page describes: a product with its price and availability, an article and its author, an organization, an FAQ. Google recommends JSON-LD. Markup can make a page eligible for rich results, but it must match what visitors actually see.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

JSON-LD (JavaScript Object Notation for Linked Data) is the format Google recommends for structured data: a script block in the page that describes it in schema.org terms, such as an organization, a product with its price or an article with its author. Because it sits apart from the visible HTML, it is easy to generate from templates and keep up to date. It must still match what the page visibly shows; markup that describes content readers cannot see breaks Google's rules, which is why we check it template by template.

More: AEO & GEO Optimization

Rich results are search results enhanced with extra elements built from structured data, such as price, availability, ratings, images or breadcrumbs; product snippets and merchant listings, for example, come from Product markup. Markup makes a page eligible but does not guarantee the display, and Google changes which types it shows: FAQ rich results stopped appearing in May 2026. Rich results often disappear after a migration when the new templates lose the structured data.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click

FAQ schema is FAQPage structured data that labels the questions and answers on a page so that search engines and AI systems can read them as pairs. Google stopped showing FAQ rich results in May 2026, so the markup no longer earns extra space in classic results. It still helps machines parse the content and has been cited as raising appearances in AI answers by 40%. FAQ blocks should therefore be written for people first, with real questions answered in full.

More: What Is AEO? Answer Engine Optimization Explained for Small Business in 2026

Product and Offer markup is structured data that describes a product and the terms of sale: name, identifiers, price, availability, variants, reviews, shipping and returns. Google uses it for product snippets and merchant listings, and AI shopping assistants read it to compare products. It must match the page and guarantees neither rankings nor rich results. In our test of 13 US online stores, only one described shipping and two described returns in their markup.

More: Product Page SEO: Descriptions, Reviews and Structured Data That Earn the Click