Book a free call

Semantic SEO

Semantic SEO organizes a website around the entities, attributes and questions its customers search for, so search engines can understand the business precisely and trust it as a source. The outcome is topical authority: each new page starts from a stronger position, and the site grows as one connected body of work rather than a pile of posts.

Book a free call Read how it works

Signs you need Semantic SEO

  • We publish every week, but organic traffic hasn’t moved in months.
  • Google keeps showing the wrong page for our main service.
  • Two or three of our pages seem to compete for the same searches.
  • Competitors with thinner sites outrank us for things we do better.
  • Every core update moves us down a few places and we never know why.
  • Our site touches lots of topics, but none of them in depth.
  • We never appear in AI Overviews or ChatGPT answers about our own services.

Results to expect

  • Google understands what your business is and what it should rank for
  • More pages ranking across the sections that drive inquiries and sales
  • One clear page for each search intent, with no internal competition
  • Content structured to be quoted in AI Overviews and AI assistants

What is semantic SEO?

Semantic SEO is the practice of building a website so search engines can understand what it means, not just which words it contains: the entities it covers, the facts it states about them, the search intent each page satisfies and how the pages relate to one another. When that meaning is clear, complete and consistent across a subject, the site earns topical authority, a search engine’s learned confidence that it is a reliable source on that subject.

I’m Qaim Raza, a semantic SEO specialist, and this discipline sits at the core of everything I do for clients. This page is the deep version: how search engines interpret meaning, the method I use to build authority, what progress looks like in the data, and the mistakes that quietly undo good work. The concepts can sound academic, so each one comes with an example from the industries I work in.

What semantic SEO is not

Semantic SEO is not a new name for old tactics, and four misunderstandings come up again and again.

  • It isn’t “LSI keywords”. Latent semantic indexing is a decades-old retrieval method, and Google’s own representatives have said there is no such thing as LSI keywords. Scattering synonyms through a page doesn’t make it semantic. Stating the right facts about the right entities does.
  • It isn’t schema markup on its own. Structured data labels entities that the visible content should already make clear. Markup on a vague page simply describes a vague page more precisely.
  • It isn’t publishing volume. Fifty loosely related posts can weaken a site that ten planned pages would have strengthened, because every unfocused URL blurs what the site is about and adds work for the search engine.
  • It isn’t a 30-day trick. Search engines build trust in a source over months of crawling, testing and observing searchers. Anything sold as instant authority is selling a shortcut that doesn’t exist.

Not sure where to start?

Book a free 30-minute call. You’ll know what’s holding your site back and what to fix first.

Book a free call

Semantic SEO vs keyword-based SEO

AspectKeyword-based SEOSemantic SEO
Starting pointA keyword list sorted by search volumeThe source context and central entity: what the site must be known for
What one page targetsOne phrase and its close variantsOne search intent and the full group of queries behind it
How relevance is shownRepeating the phrase in titles, headings and body copyStating the entity’s attributes and values in clear, checkable sentences
Internal linksMenus, tag pages and “related posts” widgetsPlanned contextual bridges with descriptive anchors that follow the map
Content depthMatched to competitors’ word countsSet by what the intent needs, plus information the ranking pages lack
How progress is judgedPositions for a list of tracked phrasesGrowth across whole sections of the topical map
Typical failureSeveral pages competing for the same intentA plan so large it never gets published

How do search engines understand meaning?

Search engines understand meaning by connecting words to entities, entities to their attributes, and queries to the intents behind them, then testing which documents satisfy those intents best. None of this is secret knowledge. It’s described in patents, research papers and Google’s own documentation, and it shows up in behavior anyone can observe in Search Console.

Entities, attributes and the knowledge graph

An entity is anything uniquely identifiable: a person, a business, a place, a product or a concept. Search engines store what they know about entities in a knowledge graph, a network of entities joined by relationships, and each fact in it can be expressed as an entity–attribute–value (EAV) triple. “Elm Street Storage (entity) offers unit sizes (attribute) from 5×5 to 10×30 (value)” is one. “Elm Street Storage (entity) has gate access hours (attribute) of 6 am to 10 pm (value)” is another.

This matters because a search engine doesn’t rank a page for a string of characters. It ranks a page as an answer about something. A page that names its entity plainly and states its attributes in simple sentences maps cleanly onto what the search engine already holds. A page that circles its subject with adjectives gives it nothing to store.

Attributes also come in layers. Some are shared by every entity of a type: every storage facility has unit sizes and an address. Some are rare: only a few facilities accept deliveries for business tenants or offer wine storage. Some are unique to one business. Covering the common attributes makes a page relevant. Covering the rare and unique ones is what makes it worth ranking above pages that only cover the basics.

Query semantics and query templates

Query semantics is how a search engine works out what a query means: which entity it’s about, which attribute is being asked for, and what shape of answer would satisfy it. “Storage unit sizes” asks for a list of values. “How big is a 10×10 storage unit” asks for one value and a way to picture it. “Storage near me open Sunday” adds a location and a constraint on access hours. That’s three queries about one entity, needing three different pages or passages.

Queries also follow query templates, repeating patterns with slots for entities. A law firm’s prospects type “[practice area] lawyer in [city]”, “how much does a [practice area] lawyer cost” and “do I need a lawyer for [problem]”. Each template implies a page type and a set of facts the answer must contain. Once you know the templates, you can plan a page for every combination worth serving instead of chasing phrases one at a time, and you can see exactly which slots you have no answer for yet.

Search engines also learn the order in which questions get asked during a task. Someone working out what a long-distance move costs is close to requesting quotes, so the page that answers the cost question should lead naturally to the one where they ask for a quote.

Retrieval, initial ranking and re-ranking

Ranking happens in stages rather than in one calculation. Retrieval pulls a set of candidate documents from the index that could answer the query. Initial ranking orders those candidates using signals that are fast to compute, such as how well the text matches, where the page sits in its site and what the search engine already knows about the site’s quality. Re-ranking then adjusts that order with more expensive evidence: how each page compares with the others on satisfying the intent, and how searchers respond to the results over time.

The staging has a practical consequence: a page can only be moved up if it was retrieved in the first place, which is why clear entity signals and a sensible place in the site’s structure come before any polishing. Google engineers have described, under oath, systems that use searcher interaction data to refine results, so pages that genuinely finish the searcher’s task tend to keep the positions they win.

Cost of retrieval

Cost of retrieval is how expensive a site is for a search engine to crawl, render, understand and rank. Every URL carries a price: fetching it, running its scripts, working out what it says, comparing it with near-duplicates and deciding where it fits. Search engines have finite resources, so a site that’s cheap to process tends to be crawled more completely and evaluated with more confidence.

The expensive patterns are predictable. Thousands of thin, near-identical pages. Templates where the main content loads behind scripts and widgets. Filter and parameter URLs generating endless variants. A moving company that creates a route page for every pair of cities, each one identical apart from the names, raises its cost of retrieval for very little return. A smaller set of route pages, each with real information about that lane such as typical distances, access problems and the busiest months, costs less to process and gives the search engine a reason to keep every one.

Knowledge-based trust and historical data

Knowledge-based trust is the idea that a source can be judged by how accurate its facts are compared with what a search engine already holds to be true. Google researchers have published a paper on estimating the trustworthiness of web sources this way. It isn’t a confirmed ranking factor by that name, but the reasoning is hard to argue with: a site that states facts precisely and consistently is a safer result to show. If a jewelry page says 14k gold is 58.3% pure gold, it agrees with the arithmetic (14 parts in 24) and with every reliable reference. If another page on the same site says 75%, the site contradicts both the wider web and itself.

Historical data is the record a search engine accumulates about a site: how its pages performed, how searchers responded, how often it published and how consistently it stayed on its subject. It’s why an established, focused site can get a new page ranking quickly, and why a site that keeps changing direction, or deletes and rebuilds its content every year, loses trust it had already earned. History can’t be faked, but it can be protected and built on.

AI search rewards the same clarity at a smaller scale: the passage. With passage ranking, Google can rank a page because one section of it answers a narrow question well, even when the page as a whole covers something broader. Every well-built section is a potential answer in its own right.

AI Overviews, AI Mode and assistants such as ChatGPT and Perplexity go a step further with query fan-out. They split a question into several related sub-queries, search for each one, and assemble an answer from the passages they retrieve, citing some of the sources. A question such as “is climate-controlled storage worth it for furniture” might fan out into sub-queries about humidity and wood, the price gap between unit types and which facilities nearby offer climate control. The sources that get pulled in are the ones with a self-contained passage answering one of those sub-queries plainly, published by an entity the system can identify. That’s why AI search optimization is built on semantic SEO rather than being a separate trick.

The semantic SEO method in depth

The method I use follows the framework I trained in through Koray Tuğberk Gübür’s Topical Authority course, refined over 6 years of hands-on SEO. Each stage below solves a specific problem created by the way search engines work, and each one sets the constraints for the next, which is why the order matters as much as the stages themselves.

  1. Source context and central entity

    Source context is the identity and purpose of a website as a source: who runs it, what it does or sells, and why it should be trusted on its subject. The central entity is the one entity the whole site revolves around, the thing every page should relate back to. Together they decide which topics belong on the site and the angle each topic is covered from.

    Two CBD websites can share a subject and still need completely different plans. A manufacturer that grows, processes and tests its own hemp has a source context built on production, so its strongest pages explain cultivation, formulation and quality control. A retailer that stocks twenty brands has a source context built on choice, so its strongest pages compare products and help buyers decide. If the retailer copies the manufacturer’s plan, it ends up writing pages it has no standing to write. Settling both definitions first stops that drift before it costs anything.

  2. Central search intent

    Central search intent is the main need that connects the central entity to the people searching for it: the task most of them are trying to complete. Every section of the site should serve that intent directly or serve a step on the way to it.

    For a self storage operator, the central search intent is roughly “find the right unit nearby and rent it with confidence”. That one sentence explains why unit guides, access details, security, pricing and online reservations all belong on the site. It also explains why a general article on decluttering a home is weak unless it leads somewhere that serves the central intent, and it tells you which pages deserve to sit closest to the homepage.

  3. The topical map: core and outer sections

    A topical map is the plan of every page a site needs to cover its subject, grouped and ordered by how closely each page serves the source context. It has two parts. The core section holds the pages closest to the central entity and to revenue: services, products, locations and the attributes buyers compare. The outer section supports the core with the wider questions around it, such as definitions, comparisons, processes and the problems people have before they know which service they need.

    The outer section is not filler. It builds topical coverage, the share of the subject’s queries your site can answer, and it gives core pages their context. A personal injury page is more convincing when the same site also explains how claims progress, what evidence matters and what a settlement negotiation involves. Order matters as well: I build the core first and plan each batch of outer pages to strengthen a specific core page. How I research and structure the plan is covered under topical map creation.

  4. Contextual hierarchy, macro and micro context, and contextual borders

    Every page has one macro context, its main topic and intent, and several micro contexts, the narrower subtopics that support it. Contextual hierarchy is the order and weighting of those contexts through the headings: the most important answer first, headings that follow the logical sequence of the question, and every subheading clearly belonging to the one above it. Together they set the page’s contextual vector, the direction its meaning travels from the title to the last paragraph. A page whose vector wanders sends a mixed signal about what it should rank for.

    A contextual border is the point where a page’s topic should stop. Take a family law firm’s divorce page. Child custody belongs on it as a micro context: a short section explaining how custody is decided during a divorce. Then the border is reached, and the full explanation of custody lives on its own page, linked from that section. Without borders, every page tries to cover everything, pages start competing with each other, and the search engine can’t tell which one to rank.

  5. Writing for information gain with clear EAV facts

    Information gain is the new, useful information a page adds beyond what the pages already ranking for its query say. A search engine has little reason to show an eleventh version of the same ten answers, so a page that only restates the consensus competes on authority alone, which a growing site rarely has enough of.

    Information gain usually comes from what only the business knows: how its own pricing works, its process, its measurements, the questions its staff hear every week and the mistakes customers make. For a jewelry store, that might be how a particular setting style affects how often stones need checking. For a storage facility, it might be which unit sizes customers moving out of a one-bedroom apartment actually rent, rather than the generic chart every competitor copies.

    The facts themselves should be written as clean EAV statements: the entity named, the attribute explicit, the value precise. “Our 10×20 units hold the contents of a typical three-bedroom house” is a fact a search engine can store and a reader can act on. “We have plenty of spacious options” is neither. Answer-first sentences, specific values and one claim per sentence make a page easy to understand, easy to quote and easy to verify, and they’re the standard behind my semantic content writing.

  6. Contextual bridges and the internal content network

    A contextual bridge is a link, or a short passage carrying a link, that connects two related contexts so both readers and search engines can follow the relationship between them. It’s the difference between a link that exists because a menu exists and a link that exists because the next question leads there.

    In a semantic site, internal links follow the topical map. Outer pages pass readers and relevance towards the core page they support. Core pages link to each other where buyers compare options. Anchors describe the target page’s main topic instead of saying “learn more”. The decluttering article from earlier earns its place with a bridge: a paragraph on what to do with belongings you aren’t ready to part with, linking to the unit-size guide. Planned this way, links stop being decoration and become a semantic content network that shows which pages matter most and how the subject fits together.

  7. The entity and technical layer

    The entity and technical layer makes the site’s identity machine-readable and its pages cheap to process. On the entity side, the business needs an entity home: one page, usually About, that states who the business is and that the rest of the web can point to. sameAs properties in its structured data connect that page to the business’s profiles elsewhere, and names, addresses and descriptions need to match everywhere they appear. That consistency is what makes a knowledge panel, the box Google shows when it’s confident about an entity, more likely over time.

    On the technical side, the aim is a lower cost of retrieval: clean crawl paths, main content rendered without waiting on scripts, canonical tags that agree with internal links, controlled faceted navigation, and Core Web Vitals inside Google’s thresholds, where INP (Interaction to Next Paint) has replaced FID as the responsiveness measure. Schema markup joins the two halves by describing the business, its services, people and products in a shared vocabulary. I go deeper into this layer on the Technical SEO page.

  8. Publishing momentum

    Publishing momentum is a steady, predictable pace of adding pages from the topical map. In practice, sites that publish on a consistent rhythm tend to be crawled more often and see new pages tested sooner than sites that publish in bursts and then go quiet.

    Momentum works best in topical batches: a complete group of related pages released together, with their links already in place, so the search engine meets a whole subtopic at once rather than fragments spread over months. Six pages on unit sizes, each linked to the others and to the main rental page, give a complete picture of that subtopic. Six unrelated posts give six isolated guesses.

  9. Measurement, refresh and pruning

    Measurement in semantic SEO happens at the level of topical map sections, and it drives two maintenance jobs: refreshing pages that are slipping and pruning pages that hold the site back.

    A refresh updates an existing page with current facts, a tighter hierarchy, new information gain and stronger bridges, while keeping its URL and the history attached to it. Pruning removes or merges pages that serve no intent, duplicate another page’s intent, or have never been shown for anything, usually by consolidating them into the strongest related page with a redirect. Both lower cost of retrieval and sharpen what the site is about. Neither is a one-off clean-up. On a growing site they run as a regular cycle, because every new batch changes which older pages matter.

What do semantic SEO results look like over time?

Semantic SEO results arrive in a recognizable order: pages get indexed, impressions spread across many queries, positions improve, and then clicks and inquiries follow. Knowing the order makes it far easier to judge whether a project is on track.

  1. Why impressions rise before clicks

    Impressions rise first because search engines test pages before they trust them. During initial ranking, a page is shown for a wide spread of related queries, often in positions few people scroll to, while evidence is gathered. Impressions climb because the page is being considered for many queries. Clicks stay low because it hasn’t yet earned the positions that receive them. As re-ranking moves the strongest pages up, clicks follow. A rising impressions line with flat clicks in the early months is usually a sign the plan is working, not failing.

  2. Which metrics show progress

    The metrics that show semantic SEO progress describe topics, not single phrases.

    • Query-to-page mapping: whether each group of related queries lands on the one page planned for it. When two URLs keep swapping for the same queries, that’s cannibalization, and the fix is merging or re-bordering pages, not writing more.
    • Section-level Search Console data: impressions, clicks and average position grouped by topical map section, using URL folders or page filters. A section that grows as a whole is gaining authority. One page spiking on its own is often noise.
    • Topical coverage: the share of the map that is published, indexed and earning impressions. When coverage stalls, it shows where to publish next.
    • Organic conversions by section: calls, form fills, reservations or sales traced back to each part of the map, so effort follows revenue.

    Tracking a handful of commercial terms still has a place, but on its own it misses most of what a semantic project changes.

How do semantic SEO and paid ads work together?

Semantic SEO and paid ads work together through one shared query map. The research behind a topical map already groups searches by entity and intent, so it doubles as a ready-made structure for Google Ads campaigns: one ad group per intent, landing pages taken from the core section, and negative keywords drawn from the queries the map deliberately leaves out.

The data flows the other way too. Search terms reports reveal which phrasings convert, and that feeds back into page priorities, titles and the wording of answers. Ads can also cover a commercial query while its page is still being tested in organic results, then step back once the organic position holds.

In restricted categories the balance shifts. Where a product can’t be advertised, or only under narrow conditions, organic search carries most of the demand, and the topical map does the planning job a campaign structure would normally do. Where social platforms do allow a category, Facebook and TikTok ads can build awareness that turns into branded searches, and those searches land on a site already built to answer them.

What are the most common semantic SEO mistakes?

The most common semantic SEO mistakes come from using the vocabulary without the discipline behind it.

  • Building a map without a source context. A topical map copied from a competitor or exported from a keyword tool describes someone else’s business, and it fills the site with pages it can’t write with any authority.
  • Treating the outer section as the whole job. Hundreds of informational articles with no bridges to commercial pages attract visitors who never meet a service.
  • Ignoring contextual borders. Long pages that try to answer every related question end up competing with the site’s own focused pages.
  • Letting facts drift between pages. Different prices, opening hours or specifications on different URLs damage trust faster than a missing detail does.
  • Pruning by traffic alone. A quiet page may be the bridge that passes relevance to a core page. Pages should be removed for serving no intent, not for being quiet.
  • Markup that disagrees with the page. Schema listing services, prices or reviews that the visible content doesn’t show confuses entity understanding and goes against Google’s structured data guidelines.
  • Scaling text without editing it. Mass-produced pages with no information gain raise cost of retrieval and risk falling under Google’s spam policy on scaled content abuse.

What you receive with Semantic SEO

DeliverableFormatWhy it matters
Source context and central entity briefGoogle Doc

Sets the boundaries every page, link and claim is checked against.

Topical mapGoogle Sheet + Figma board

Shows the whole plan at a glance and the order it will be built in.

Query-to-page mapGoogle Sheet

Assigns each query group to one URL, so your pages stop competing.

Content briefsGoogle Docs

Let any writer, mine or yours, produce pages that fit the map.

Written and optimized pagesLive pages in your CMS

Published with headings, facts, links and markup already in place.

Internal linking planGoogle Sheet

Source page, target page and anchor for every planned link.

Schema markupJSON-LD in your theme

Gives search engines a machine-readable version of who you are and what you offer.

Monthly performance reportLooker Studio dashboard + Loom video

Shows progress by topical map section and what happens next.

How Semantic SEO works, step by step

  1. Discovery and source context

    Week 1

    We define your source context, central entity and central search intent together. Nothing is planned or written until these are agreed in writing.

  2. Site, query and competitor analysis

    Weeks 1–2

    I crawl the site, export Search Console query data and study how competing sites cover the subject, so the map starts from evidence rather than guesswork.

  3. Topical map and query-to-page map

    Weeks 2–4

    The full map is built and prioritized. We review it together on a call, and you approve the first batch before any writing starts.

  4. Technical and entity fixes

    Weeks 3–5

    Crawl, indexing and template problems are fixed, the entity home is set up and core schema goes live. This step needs CMS or developer access.

  5. Briefs and first content batch

    Weeks 4–8

    Briefs for the first batch of core pages are written, then the pages are drafted by me or by your team and edited against the brief.

  6. Internal links and publishing

    Weeks 6–12

    Each batch goes live with its links in place, and older pages are updated to bridge into the new ones.

  7. Measure, refresh and expand

    Week 12 onwards

    Monthly reporting, refresh and pruning cycles, and new sections of the map in the order the data supports.

Semantic SEO in the industries I specialize in

Kratom

Kratom sites need a topical map that builds trust as carefully as it builds traffic. The core section covers strains, vein colors and formats such as powder and capsules, described by origin, processing and lab results rather than by effects. Outer sections answer legality by state and city, shipping restrictions, age verification and how to read a certificate of analysis. Because the FDA hasn’t approved kratom for any use, no page makes health or medical claims.

CBD

For CBD brands, trust content belongs in the core of the map, not in a footer: batch-level certificates of analysis, hemp sourcing, hemp-derived status and THC limits under federal and state law each get a clear, current page. Product and category copy describes cannabinoid content, format and serving size, and makes no disease or cure claims, the wording FDA warning letters target. Topical and ingestible products sit in separate sections because their rules differ.

Movers

Moving is a high-trust purchase in a category known for scams, so the map gives trust its own section: how to verify a USDOT number with the FMCSA, what separates a broker from a carrier, and how binding and non-binding estimates differ. Local and long-distance services are planned as separate core sections. Because Google Business Profile spam and suspensions are common for movers, the site also has to prove the business is real, with license details and verifiable reviews.

Self storage

Self storage operators compete with REIT brands and aggregator sites for “storage units near me”, and semantic SEO is how an independent operator wins back ground. Each facility page carries local details a listing can’t match: directions, nearby landmarks, vehicle, RV and boat storage where offered, and a direct reservation path. Multi-location operators get one location template, so every facility page is consistent but never a copy of the others.

Jewelry stores

Jewelry is the most attribute-rich niche I work in. Engagement ring categories are planned around metal, karat, cut, carat weight, color, clarity and certification such as GIA or IGI, with lab-grown and natural stones described in line with the FTC Jewelry Guides. Stores that also sell online need two connected structures: local pages for custom design, repairs and appointments, and e-commerce categories that support them rather than compete with them.

Law firms

Law firm sites are YMYL, so the entity layer carries as much weight as the writing. Each attorney gets a Person page with bar admissions, education and practice areas, connected to the firm through LegalService and Person schema. Practice-area pages are planned per jurisdiction, because procedure and deadlines differ by state. All of it follows state bar advertising rules based on ABA Model Rule 7.1: nothing misleading, no outcome guarantees, and no “specialist” claims where a state restricts them.

Ways we can work together

Option 1

Topical map and briefs

Best for: Businesses with writers who need a plan they can follow

  • Source context and central entity brief
  • Full topical map and query-to-page map
  • Briefs for the first batch of pages
  • Review of your team’s first drafts
Option 2

Done-for-you semantic SEO

Best for: Owners who want the whole system built and run for them

  • Everything in the map and briefs option
  • Writing, publishing and internal linking
  • Technical fixes, entity home and schema
  • Monthly reporting and refresh cycles
Option 3

Monthly retainer

Best for: Sites with a map in place that need steady expansion

  • New pages each month in map order
  • Refresh and pruning of existing pages
  • Ongoing technical and schema upkeep
  • Monthly report and next-month plan

Tools I use

  • Google Search Console
  • Google Analytics 4
  • Screaming Frog SEO Spider
  • Ahrefs
  • Semrush
  • Google Cloud Natural Language API
  • Looker Studio
  • Schema Markup Validator

Who this is for, and who it isn’t

A good fit if you…

  • Businesses with real expertise and information competitors can’t copy
  • Owners ready to publish steadily for at least six months
  • Businesses willing to share their own data, prices and process
  • Sites that have published plenty but lack structure
  • Multi-location operators that need one system across every location

Not the right fit if you…

  • You need sales this week: paid ads are the faster channel for that
  • You want guaranteed rankings or a fixed number of backlinks
  • You want hundreds of pages published quickly, whatever the quality
  • Your site can’t be edited, or every page needs months of sign-off
  • You want claims on the site that regulators or bar rules don’t allow

Semantic SEO questions

What is semantic SEO in simple terms?

It’s organizing your website the way a good reference book is organized. A reference book has one clear subject, covers each subtopic once and in a sensible order, states facts precisely and points you to related chapters when you need them. Semantic SEO gives a website those same qualities, so search engines can tell what it’s about, which page answers which question, and why it deserves to be trusted.

How is semantic SEO different from keyword SEO?

The main difference is the unit of work: keyword SEO plans around phrases, while semantic SEO plans around intents and entities. A keyword approach might create separate pages for “cheap storage units”, “affordable storage units” and “low-cost storage”, which then compete with each other. A semantic approach recognizes one intent, serves it with one strong page, and spends the saved effort on questions competitors haven’t answered.

What is a topical map, and do I need one?

A topical map is the full list of pages your site needs, each with its intent, its section and its place in the publishing order. You need one if you plan to publish more than a handful of pages, if several of your pages compete for the same searches, or if content has been added for years without a plan. A small site with one service may only need a short version.

What are source context and central entity?

They’re the two decisions everything else depends on: what kind of source your site is, and the one thing it revolves around. For a jewelry store that designs engagement rings in its own workshop, the central entity is the store and its design work, and the source context is a maker rather than a reseller. That puts design process, materials and workshop pages at the center of the map and keeps celebrity ring news out.

What is topical authority, and how do you measure it?

Topical authority is a search engine’s confidence that your site is a reliable source on a subject. No tool reports it directly, so I measure its effects. The clearest is speed: when new pages in a section start earning impressions and settle into stable positions faster than the first pages in that section did, the site has earned trust on that topic. Ranking for useful queries you never targeted is another good sign.

How long does semantic SEO take to show results?

Impressions usually start growing within weeks of the first planned pages being indexed. Meaningful clicks typically take three to six months, and well-established topical authority six to twelve, depending on competition, the site’s history and how consistently the map is published. YMYL and restricted niches tend to sit at the slower end. After reviewing your site I’ll give you an honest range, not a guarantee.

Why are my impressions growing but clicks are not?

Early in a project, that pattern is expected and usually healthy. If it lasts more than a few months, I check three things. Do the impressions come from queries the page should own, or from loosely related ones? Do the title and snippet match the intent behind those queries? And is an AI Overview, map pack or shopping block now answering the query above the organic results?

How many pages will my site need?

As many as the subject requires, and no more. The topical map sets the number by counting the intents worth serving, not by hitting a word count or a publishing quota. A single-location service business often needs dozens of pages, while a multi-location operator or a firm with several practice areas can need hundreds. Fewer, stronger pages almost always beat more, thinner ones.

Will you rewrite my existing content or start fresh?

Usually I improve what you have. Every existing URL is matched to the new topical map first, which shows what already fits, what overlaps and what has no place. Pages with history are the cheapest wins because search engines already know them. Starting fresh only makes sense when a site has almost no indexed pages, or when its history is damaged, for example by a past manual action.

What is information gain, and why does Google reward it?

Information gain is what a page adds that the pages already ranking don’t. A Google patent, “Contextual estimation of link information gain”, describes scoring documents by how much new information they offer beyond what a searcher has already seen. Whether or not that exact system is in use, the logic holds: showing ten versions of the same answer helps nobody, so search engines are designed to prefer the page that adds something.

Does semantic SEO help me appear in Google AI Overviews and ChatGPT?

It improves your chances, though nobody can guarantee a citation. Google says pages shown as links in AI Overviews and AI Mode need to be indexed and eligible for a snippet, with no special markup required, so solid organic foundations come first. ChatGPT and Perplexity run their own web searches and cite pages they can retrieve and understand. I track a fixed set of prompts each month to see where your brand appears and who is cited instead.

Do you use AI to write content? Is AI content safe?

I use AI tools where they genuinely help: clustering queries, summarizing research and checking drafts for gaps. Every page is written or heavily edited by a person against a brief, with facts from you and from primary sources. Google’s guidance judges content on whether it’s helpful and reliable, not on how it was produced. The real risk is publishing unedited text that says nothing new and gets details wrong.

Do I still need backlinks?

Usually yes, but often fewer than you’d expect, and of a different kind. A site with clear topical authority can outrank competitors that built links to weak pages. Links still matter in competitive local markets and YMYL niches, and the most useful ones come from sites relevant to your subject that mention your brand in context. In CBD and kratom many publishers refuse the topic, so expectations have to be realistic.

How does semantic SEO work for local businesses like movers, self storage and law firms?

Local semantic SEO treats each location as an entity with its own facts and ties it to the matching Google Business Profile. Profile categories and services mirror the site’s service pages, each location page carries details only that branch has, and reviews, citations and address details agree everywhere. These businesses compete in the map pack as well as the organic results, so both are planned together.

Can semantic SEO work for CBD and kratom brands where ads are restricted?

Yes, and for many of these brands it’s the main channel. Google Ads, Meta and TikTok don’t allow kratom ads, and ingestible CBD is heavily restricted on all three. Google permits limited ads for topical hemp-derived CBD in a few US states with LegitScript certification, under rules that change, so I check the current policy before any launch. Organic search has no such ban, which makes a compliant, well-structured site the most dependable route to new customers.

How do you handle YMYL topics such as legal or health claims?

With stricter checks at every stage. Claims are sourced from primary references such as statutes, court rules, regulator guidance or lab reports. A qualified person on your side reviews each page before it goes live, and pages show who wrote and who reviewed them. Wording is conservative by default: if a statement can’t be supported by a source, it doesn’t go on the page, however well it might rank.

Can my in-house writers use your briefs?

Yes. The briefs are written for writers who don’t know semantic SEO. Each one sets the page’s macro context, heading order, the facts to state, the information gain angle, the questions to answer, the contextual border and the internal links with their anchors. I review the first drafts against the brief and give written feedback, so your team learns the method rather than just following a template.

What do you need from me to start?

Access and some of your time. I need Search Console, Google Analytics 4 and CMS access, plus Google Business Profile access for local businesses. In our first session I ask what you sell, who buys, what makes you different and what must never be said about your business. After that, expect roughly an hour a week for reviews and questions.

What do I own when the project ends?

Everything. The source context document, topical map, briefs, pages, schema and dashboards are created in your accounts or handed over in editable formats, and they stay yours whether or not we keep working together. There’s no proprietary platform and no lock-in. A well-documented topical map also means another writer or agency could continue the work without starting over.

How do you report progress?

Monthly, in plain English. Each report covers what changed, why I think it changed and what I’ll do next, with the numbers behind each point. Between reports you can email me. You work with me directly from start to finish, with no account managers or handoffs, so the person explaining the results is the person doing the work.

What happens after the topical map is delivered?

The map becomes a working document rather than a finished one. Batches are built from it in order, and it’s updated whenever Search Console shows queries we didn’t plan for, when you add services or locations, or when competitors cover ground we left open. A map that hasn’t changed after six months usually means nobody is reading the data.

Let’s make Google see you as the expert.

Book a free call, or message me directly. I reply within one working day.

Book a free call
Email[email protected]