People find answers in classic results, in AI Overviews, in chat assistants and through voice. Modern SEO covers all of them — three connected layers of the same job.

Think of it likeThink of a city guidebook: SEO gets you listed, GEO gets the tour guide to recommend you by name, AEO gets your words quoted.

Sub-topics

What is SEO

Improving a site so search engines can find it, understand it and show it for relevant searches.

What is GEO

Making content visible inside AI-generated answers (ChatGPT, AI Overviews, Perplexity) by targeting what AI retrieves.

What is AEO

Structuring content so answer systems — snippets, voice, chat — can lift a clean, direct answer out of it.

Bots constantly crawl the web by following links, index what they find in a giant database and — when you search — retrieve and rank the most helpful pages in an instant.

Think of it likeA librarian for the entire internet: walks the shelves, catalogues every book, then hands you the best ones.

Sub-topics

Crawling, Indexing, Ranking

The 3 core jobs. Crawling finds pages, indexing files them, ranking orders them for each query.

Retrieval vs Ranking

Two separate steps. Retrieval grabs a shortlist of candidates; ranking sorts that shortlist.

Personalization & Localization

Results change with who searches (history, device) and where (location, language).

Training Data vs Grounding

Training data = what an AI model memorised (frozen). Grounding = it looks up live web pages while answering.

The words you will hear every day. Learn them once and the rest of the roadmap reads easily.

Think of it likeCrawl = read it · Index = file it · SERP = show it.

Sub-topics

SERP, Index, Crawl

SERP: the Search Engine Results Page. Index: the stored database of pages an engine can show. Crawl: a bot visiting and reading a page — it must happen before indexing.

Backlinks & Anchor Text

A backlink is a link from another site to yours — treated like a vote of trust. Anchor text is the clickable words; it tells engines what your page is about.

Canonical

An HTML tag naming the master version of a page when several URLs show the same content. Ranking signals get combined, not split.

Nofollow, Noindex, Disallow

Three different “stop” signs: nofollow = don’t pass trust through this link; noindex = crawl it but don’t show it in results; Disallow (robots.txt) = don’t crawl this path at all.

Organic vs Paid vs Direct

Organic: unpaid search results earned through SEO. Paid: search ads labelled “Sponsored”. Direct: someone typed your URL or used a bookmark.

Understand what people mean, not just what they type — then match the page to that need.

Think of it likeSearch the keyword, study the top 10, build the winning format — but better.

Sub-topics

Search Intent

The goal behind a search: to learn, to go somewhere, to compare, or to buy. Matching intent matters more than matching the exact keyword.

Semantic Search

Finding content by meaning, not exact words — so “cheap flights” also finds “affordable plane tickets”. It’s what lets modern search and AI understand paraphrases.

Lexical vs Vector Search

Lexical search matches the literal words (like Ctrl+F). Vector search turns text into numbers and matches meaning, so “cheap flights” also finds “affordable plane tickets”.

Hybrid Search

Runs lexical and vector search together and merges the results — catching exact names and codes as well as paraphrases. Most modern search and AI retrieval works this way.

Three “styles” of SEO, defined by how they treat search engine guidelines: white hat improves the site for users, grey hat bends the rules, black hat deliberately breaks them to manipulate rankings.

Think of it likeLike a diet: white hat = eat well & exercise; black hat = crash diet pills.

The published rulebook (Google Search Essentials) listing manipulative practices. Breaking it can trigger a manual action or an automatic demotion.

Think of it likeJudge every tactic by one question: does this help the visitor?

Sub-topics

Link Schemes

Links built to game rankings instead of earned on merit — buying links, excessive link swaps, automated link networks.

Cloaking & Doorways

Cloaking shows bots different content than people. Doorways are thin pages made only to rank and funnel visitors elsewhere.

Scaled Content Abuse

Pumping out lots of low-value pages — by AI or humans — mainly to rank, with little oversight or real help for readers.

Site Reputation Abuse

Third-party, low-value content hosted on a trusted site to borrow its authority (“parasite SEO”).

Expired Domain Abuse

Buying an old domain for its history, then filling it with unrelated content to ride its past reputation.

Do this

• Read Google’s spam policies once — they’re short
• Mark paid or affiliate links rel="sponsored"
• Judge every tactic by: does this help the visitor?

A manual action is a penalty applied by a human reviewer for breaking guidelines — it can demote or remove pages or the whole site. Recovery means fixing every instance of the cause, then submitting a reconsideration request.

Think of it likeYou get told exactly what is wrong — fix it fully before asking for a review.

Do this

• Check Search Console → Manual actions
• Fix and document every instance
• Send a reconsideration request; reviews take days to weeks

Changes to how search engines rank pages. They can shift rankings across the web without any site being penalised.

Think of it likeLike a school changing its grading rubric — your essay didn’t change, the scoring did.

Sub-topics

Core Updates

Broad changes to the main ranking systems, a few times a year. No single fix — improve real quality & relevance over time.

Spam Updates

Narrow updates aimed at sites using manipulative tactics. Only spammers should feel them.

Helpful Content System

Rewards content made for people, demotes content made mainly to rank. Since 2024 it’s part of Google’s core ranking systems.

Tracking Volatility

Watching sudden ranking swings across your site or industry to tell an update apart from a site-specific problem.

Do this

• Annotate update dates in your analytics
• Wait for the rollout to finish before reacting
• Fix weaknesses, don’t chase every wobble

How pages are organised and connected. Users and crawlers both use this structure to navigate and to understand what matters most.

Think of it likeA supermarket: aisles (folders), signs (URLs), and short walks to the popular stuff (shallow depth).

Sub-topics

Folders & URL Structure

Group pages in logical folders (/blog/, /products/). Readable URLs tell users and bots what a page is before it loads.

Internal Linking

Links between your own pages. They help discovery, share authority, and signal what’s related. More internal links = “this page matters”.

Site Hierarchy & Depth

Homepage → categories → pages. Depth = clicks from the homepage. Flatter = easier to crawl fully.

Machine-Readable Content

Real HTML text and headings — not key info hidden in images, PDFs-only or scripts — so bots and AI can extract it.

Sitemaps

An XML list of your important URLs, submitted in Search Console. Helps discovery; doesn’t guarantee indexing.

Robots.txt

A text file at /robots.txt telling bots which paths to skip. Polite bots obey; it’s not a security tool.

Every time a browser or bot asks for a URL, the server answers with a 3-digit code: 2xx works · 3xx moved · 4xx missing · 5xx broken.

Think of it likeLike the post office: delivered, moved, no such address, closed.

Sub-topics

Redirect Chains & Loops

A chain is several redirects in a row before the final page — slower and wasteful for crawlers. A loop points back to an earlier URL, so the page never loads. Redirect straight to the final URL.

Crawling = a bot visits and reads a page. Indexing = the engine stores and organises it so it can be retrieved later. A page can be crawled but not indexed — e.g. if it’s judged thin or a duplicate.

Think of it likeCrawled ≠ indexed ≠ ranking.

Sub-topics

Crawl Budget

How much crawling an engine will spend on your site in a period. Big or messy sites burn it on junk URLs, so key pages get visited less.

Index Coverage & Bloat

Coverage = how much of your site is actually in the index. Bloat = lots of thin, duplicate or low-value pages indexed, diluting your site’s quality.

Log File Analysis

Reading your raw server logs to see what bots really did — which pages they hit, how often, and which they ignore.

Managing AI Crawlers

Choose how AI systems may use your content — balancing visibility in AI answers against control of your data.

Controls & Pay-Per-Crawl

Network-level controls (e.g. Cloudflare) can block known AI bots by default. Pay-per-crawl lets a site charge AI companies for access instead of choosing only “free” or “blocked”.

The bots reading your site: search crawlers feed classic indexes, AI crawlers collect content for training or AI search, and site crawlers are audit tools you run yourself.

Think of it likeDecide who gets in — and for what.

Sub-topics

Googlebot

Discovers and reads pages for Google’s index. Renders JavaScript fairly well, but slow or heavy scripts can still hide content.

Bingbot

Crawls for Bing — whose index also powers other services, so being invisible to Bingbot costs more than Bing traffic.

GPTBot

Crawls content that may train OpenAI models. ChatGPT search uses a different bot, OAI-SearchBot.

ClaudeBot

Collects public pages that may train Claude models. Claude-SearchBot & Claude-User handle search and user requests.

Google-Extended

Not a separate bot — a robots.txt token controlling whether your content helps train & ground Gemini. Doesn’t affect Google Search ranking.

PerplexityBot

Indexes pages so Perplexity can find and cite them in answers.

Screaming Frog tool

Simulates a search bot on your site: broken links, missing titles, duplicates, blocked pages. The industry workhorse.

Sitebulb tool

Similar audit crawler with more visual, prioritised reports and plain-English explanations.

If content is built by JavaScript in the browser and a crawler can’t run it properly, the crawler sees a blank or half-empty page. Goal: what bots receive = what visitors see.

Sub-topics

SSR vs CSR vs Hydration

SSR sends finished HTML, so bots see everything. CSR builds the page with JavaScript in the browser — empty until it runs. Hydration sends HTML first, then JavaScript adds interactivity.

Canonicalization

Tell engines which URL is official when the same content lives at several. Self-canonical every page; point variants to the main one.

Pagination

Content split over page 1, 2, 3… Give each page its own URL & self-canonical, link them with normal links. Don’t canonical page 2 to page 1.

Faceted Navigation

Filters (size, colour, price) can create millions of near-duplicate URLs. Index only valuable combos (“red trail shoes”); keep the rest out.

URL Parameters

Bits after? — sort orders, tracking codes. They multiply URLs for the same content; canonicalise to the clean URL.

Do this

• Put key content & links in the initial HTML (SSR / static)
• Use real <a href> links, not JS click handlers
• Check “View crawled page” in Search Console URL Inspection

Any big change to a site’s domain, structure or platform — a new domain, HTTP→HTTPS, redesigned URLs, a new CMS. Done carelessly it can wipe out years of rankings.

Think of it likeMoving house: redirects are your forwarding address.

Sub-topics

Planning & Checklists

Document the current site (every URL, title, ranking page), decide what changes, and build a checklist so nothing slips through.

Domain & Protocol Changes

New domain, or switching http→https / www→non-www. Both need sitewide 301s. For domain moves use Search Console’s Change of Address tool.

Redirect Mapping

A complete old→new list. Unmapped URLs look “gone”, not moved.

Post-Migration Monitoring

Watch rankings, traffic, 404s, redirects and crawl stats closely for weeks after launch — and fix problems before they stick.

Do this

• Plan → map → launch → monitor
• One-to-one 301 redirects
• Check robots.txt and noindex on launch day

Google’s user-experience metrics: loading, responsiveness, visual stability. One ranking signal among many — but slow pages lose users either way.

Think of it likeMeasured on real users, at the 75th percentile.

Sub-topics

PageSpeed Insights

Free Google tool: shows field data (real Chrome users) + lab data (Lighthouse test) and specific fixes.

Do this

• LCP ≤ 2.5 s (loading)
• INP ≤ 200 ms (responsiveness)
• CLS ≤ 0.1 (visual stability)

Google ranks using your mobile version (mobile-first indexing).

Do this

• Responsive layout, no sideways scroll
• Readable text, tappable buttons
• Same content as desktop

Encrypted connection = the padlock. A baseline trust signal, not a big ranking boost — but its absence scares users away.

Do this

• 301 all http → https
• Fix mixed-content warnings
• Watch GSC Security Issues

On-page SEO = optimising each page’s content and HTML so it ranks better and earns more relevant clicks.

Think of it likeTune both the content and the HTML of every page.

Sub-topics

Title Tags

The clickable headline in results & tabs. ~50–60 characters, main topic first, unique per page. Biggest on-page signal & drives clicks.

Meta Descriptions

The ~150–160 character summary under the title. Doesn’t affect ranking, but a good one wins the click. Google may rewrite it.

Heading Structure

One H1 for the topic, H2s for sections, H3s inside them. Helps scanning and shows how ideas relate.

URL Slugs

Short, lowercase, hyphenated, descriptive: /best-trail-shoes/ — not /p?id=8812.

Internal Anchor Text

Link with words that describe the target (“trail shoe sizing guide”), never “click here”.

Image Optimization

Descriptive file names & alt text, compressed WebP/AVIF, right dimensions, lazy-load below the fold.

Formatting & Readability

Short paragraphs, bullets, tables, plain words. Easier for readers — and for AI to extract.

Keyword Placement

Use the main phrase in title, H1, intro and a few headings — naturally. Stuffing reads badly and can hurt.

Finding the actual words people search, then judging which are worth a page.

Think of it likeOne page per intent, many pages per topic.

Sub-topics

Volume, Difficulty & Potential

Volume: est. monthly searches. Difficulty: how hard to crack the top 10. Traffic potential: realistic visits incl. related terms one page can rank for.

SERP Analysis

Study what already ranks: content type, format, angle, and features (snippets, videos, AI Overview). Copy the intent, beat the quality.

Semrush tool

All-in-one SEO platform: keyword research, rank tracking, competitor and backlink analysis, plus an AI visibility toolkit.

Ahrefs tool

SEO platform with one of the largest independent backlink indexes — strong for keyword and link research.

Google Keyword Planner tool

Free Google Ads tool with search volume estimates straight from Google’s data.

A pillar page covering a broad topic, linked to supporting pages that go deep on each sub-topic. Shows search engines and AI that you cover the subject completely.

Think of it likeLike a textbook: one chapter overview, many detailed sections — all cross-referenced.

Do this

• One main keyword group per page — avoid two pages fighting (cannibalisation)
• Link every support page up to the pillar, and the pillar down to each

Find the real questions people ask and build content around them — richer context than keyword tools alone.

Sub-topics

People Also Ask

Google’s expandable related questions. A goldmine of real questions — and a feature you can appear in.

Forums & Communities

Reddit, Quora & niche forums: candid questions in people’s own words — and sources AI often pulls from.

AI Chat Prompts

How people phrase requests to ChatGPT & co. — longer and more conversational than search keywords.

Find the topics, keywords and formats your competitors rank for but you don’t.

Do this

• Pick 3 real search competitors (not just business rivals)
• Sort gaps by intent fit & potential
• Build better pages — don’t clone theirs

Google’s framework for judging content quality & credibility. Not a single ranking factor or score — a lens. AI systems use similar signals when picking sources.

Do this

• Real author names, bios & credentials
• Show your work: photos, tests, sources
• Clear About, Contact & editorial policy

Does your content add something genuinely new — a fact, data point, example or viewpoint — compared to what already ranks? If it only repeats what’s out there, search engines and AI have no reason to prefer it.

Think of it likeThe 11th copy of the same article has no reason to rank.

Original research, your own data, and custom-made images and charts — instead of stock photos and numbers borrowed from someone else’s report. The strongest form of information gain, because nobody else has it.

Think of it likeBe the source others quote — not the one quoting others.

Sub-topics

Original Research & Surveys

Run your own study. e.g. survey 500 runners about injuries. New data = a reason for journalists and AI to cite you.

Proprietary Datasets

Data only you have — product usage, customer or sales trends. e.g. “Our app users ran 12% further this year.” Hard to copy.

Case Studies & Benchmarks

Real outcomes and comparisons. e.g. “How we cut LCP from 4.1 s to 1.9 s” or “average CTR by position in our niche”.

Custom Charts & Images

Visuals built from your own data — a unique asset tied to your page, unlike stock images found everywhere.

Do this

• Publish the method (sample size, dates) so it’s citable
• Put the key stat in a sentence, not only in an image

Optimising for how search and AI understand distinct real-world things — people, companies, products, places — and how they connect. Goal: systems correctly identify what you are, how you relate to others, and that all mentions of you are the same you.

Think of it like“Nike” the company ≠ “nike” the word.

Sub-topics

Entities vs Keywords

A keyword is text (“nike shoes”). An entity is the actual thing behind it — Nike, the company, with a history and HQ.

Attributes & Relationships

Attributes = facts about an entity (founded 1964, sportswear). Relationships = how entities connect (Phil Knight → co-founded → Nike).

Disambiguation & Salience

Disambiguation: which “Michael Jordan” is this? Salience: how central an entity is to the page — about it vs. mentioning it.

Knowledge Graphs

Databases of entities & relationships. Google runs its own; Wikidata is a big open one many systems use. Being in them = being recognised.

Code (usually JSON-LD) that explicitly describes what’s on a page — this is a product, this is the author, this is the organisation. Schema.org is the shared vocabulary. It can unlock rich results and removes guesswork for search & AI.

Think of it likeLike name tags at a conference.

Sub-topics

Article, FAQ & HowTo

Article: headline, author, dates. FAQPage: question-and-answer lists. HowTo: step-by-step instructions. (Google retired HowTo rich results in 2023 and stopped showing FAQ rich results in May 2026; the markup is still harmless.)

Breadcrumb, Video & Review

Breadcrumb: Home › Blog › Post. Video: thumbnail, duration, key moments. Review: ratings that can show as stars.

SameAs & Entity Linking

sameAs points to your official profiles (Wikipedia, Wikidata, LinkedIn) — “these are all me”.

Org. & Person Schema

Name, logo, founder, job title, affiliation — clear facts about who is behind the site.

Product & Place Schema

Product: price, availability, ratings. Place / LocalBusiness: address, opening hours, location.

Authors & Brands

Mark who wrote it (Person + bio URL) and which brand makes it — supports trust and entity recognition.

Format content so an AI system can pull a clean, direct answer without decoding long, rambling paragraphs.

Think of it likeAnswer first, explain second.

Sub-topics

TL;DRs

A short summary at the top stating the main point — a ready-made answer for readers and AI.

Data Tables

Rows & columns keep facts separate, so they’re easy to scan and extract accurately.

Direct Answers

State the answer plainly, right under the question — don’t bury it in paragraph four.

The highlighted answer box at the top of Google, lifted from a page in the results. You win it with a clear, well-formatted answer near the top of the page.

Think of it likeYou’re not writing for a reader’s eyes only — you’re writing a quote that can be cut out.

Do this

• Target questions you already rank top-10 for
• Match the format the current snippet uses — then answer better
• Keep the answer concise; go deeper underneath

Non-standard elements on a results page. Appearing in them can change your visibility and clicks far more than moving one position.

Think of it likeEach box on the results page is a different way to be seen.

Sub-topics

Sitelinks

Extra links under a result to your key pages. Automatic for clearly structured sites — you can’t request them.

Knowledge Panels

A fact box about an entity, usually on the right, fed by knowledge graphs. Needs consistent entity information everywhere.

Image & Video Packs

Rows of thumbnails for visual queries. Needs good image/video optimisation and structured data.

Breadcrumb Rich Result

Shows the page’s path (Home › Guides › Shoes) instead of the raw URL. Powered by breadcrumb schema.

Everything after publishing: monitor performance, then decide to refresh, rewrite, merge or remove — and formally de-index what no longer serves a purpose.

Think of it likeGardening: plant, water, prune.

Sub-topics

Content Freshness & Decay

Freshness: does it still read as current? Decay: slow loss of rankings as it ages and rivals improve.

Duplicate Content

Same or near-same content on several URLs confuses ranking and splits signals.

Content Consolidation

Merge weak, overlapping pages into one strong resource; 301 the old URLs to it.

Pruning & Deindexing

Prune: remove/merge thin or outdated pages. Deindex: tell engines to drop a page (noindex, 404/410).

Platforms that analyse top-ranking pages for a topic and suggest what your draft is missing — sub-topics, terms, depth.

Sub-topics

Surfer SEO tool

Suggests related terms, length & structure for a draft based on what ranks.

Clearscope tool

Grades your writing against top content and lists missing concepts.

MarketMuse tool

AI content planning: finds gaps across a whole topic to build authority.

Frase tool

Keyword research + content briefs + AI writing help in one place.

More people ask questions in chat interfaces instead of typing keywords. Content now serves two audiences: classic crawlers and AI systems that retrieve and summarise.

Think of it likeContent now serves two audiences: crawlers and AI.

How an AI answer gets built: the question fans out into sub-queries, matching passages are retrieved and re-ranked, then a model writes the answer and cites its sources.

Think of it likeYour content has to win at the “retrieve” step — or it can’t be cited.

Sub-topics

RAG

The AI first retrieves relevant passages from an index or the live web, then generates its answer from them. That’s why AI answers can cite current pages.

Embeddings

Text turned into lists of numbers that capture meaning. Similar meanings land close together — even with zero shared words.

Chunking

Long pages are cut into smaller passages before indexing. Systems retrieve chunks, not whole pages — so each section must stand on its own.

Dense Retrieval

Finding passages by comparing the question’s embedding with stored embeddings — matching by meaning (vector search).

Query Fan-Out

One question is split into several hidden sub-queries (pricing, integrations, reviews…). You can be retrieved for a piece of the answer.

Multi-Step AI Intent

Complex questions trigger several rounds: retrieve → check what’s missing → retrieve again → answer. Common in chat & agent modes.

Where AI answers appear. Each surface picks its sources differently, so check your key questions in several of them.

Think of it likeBe present where the AI looks.

Sub-topics

AI Overviews

Google’s AI summary above the results. Built from indexed pages, so normal SEO quality still decides who is cited.

Google AI Mode

A conversational Google search with follow-up questions; goes deeper than an Overview and also draws on retrieved web content.

ChatGPT Search

ChatGPT fetches live web results and answers conversationally with linked sources.

Perplexity

An answer engine: searches the web and returns a synthesised answer with numbered citations.

Copilot & Gemini

Microsoft’s assistant (built into Bing) and Google’s assistant — both retrieve web information to answer.

Chat-Based Search

Asking an assistant like ChatGPT or Claude instead of a search box — answers come from the web, training data, or both.

The sites AI systems pull from most when answering — community forums, encyclopedias, video platforms, reviews, press. Knowing them tells you where to have a genuine presence.

Think of it likeBe present where the AI looks.

Within a topic, each AI system tends to favour the same handful of sites — shaped by authority, clear structure and crawl access.

Do this

• Run your top 20 questions in each AI tool
• List the domains cited again and again
• Earn a place there: PR, guest expertise, community answers

An AI that takes actions for the user — browsing several pages, comparing options, then booking or buying. Search becomes one step in finishing a task.

Think of it likeMake your site easy for robots to use, not just read.

Sub-topics

How Agents Browse

Agents read page content, follow links and interpret forms & buttons on their own. Clear HTML, labelled buttons and simple forms help them succeed.

MCP & Structured Access

Model Context Protocol: a standard way for AI to plug into tools & data directly — like a USB-C port for AI — more reliable than scraping pages.

Getting other websites to link to yours. Search engines read relevant, high-quality links as trust and authority — a page with good links is treated as more credible than one with none.

Think of it likeLike references on a job application.

Sub-topics

Domain Authority

Moz’s 1–100 estimate of how likely a whole site is to rank, based on its links. Useful for comparing sites — not a Google metric.

Page Authority

The same Moz idea for a single page, based on the links pointing to it.

Trust Flow

Majestic’s measure of link quality: how close you are to trusted sites.

Citation Flow

Majestic’s measure of link quantity, regardless of quality. Citation Flow far above Trust Flow hints at low-quality links.

Nofollow, Sponsored, UGC

Link attributes: nofollow = don’t pass credit; sponsored = paid, ad or affiliate link; ugc = posted by users in comments or forums.

Ahrefs tool

Backlink data: your link profile and competitors’ links.

Majestic tool

Backlink analysis behind Trust Flow and Citation Flow.

BuzzStream tool

Outreach CRM for link building and digital PR.

Pitchbox tool

Outreach automation for large link-building campaigns.

Earning media coverage from real publications by pitching newsworthy stories, data or angles — instead of asking for a link.

Think of it likePitch stories, not link requests.

Sub-topics

Data-Driven PR

Run a survey or study, then pitch the stats as a story. e.g. “62% of runners skip warm-ups” → fitness & health press. Concrete numbers are easy for journalists to cite.

Getting individual creators, bloggers or YouTubers to mention, review or link to you. Like digital PR, but aimed at people with their own audiences rather than publications.

Think of it likeA trusted friend’s recommendation beats a billboard.

Do this

• Choose creators your audience already trusts
• Offer real value: early access, data, expertise
• Paid or gifted? Disclose it and use rel="sponsored"

How uniformly your brand or person appears across the web — including places you don’t control. Search and AI use this to confirm that scattered mentions all refer to one real entity.

Think of it likeConsistency is how machines confirm identity.

Sub-topics

Author & Org. Consistency

Same name, bio, job title and facts on your site, guest posts and profiles. Mismatches (“Head of SEO” vs “SEO Lead”, typos) blur identity.

Directories & Databases

Industry directories and structured databases act as independent confirmation that you exist and that your details are right.

Social Profiles

LinkedIn, X and similar profiles are cross-checked by knowledge graphs when verifying who or what you are.

Unlinked Brand Mentions

Your name mentioned without a link. Little classic link value — but real evidence that you exist and are discussed, which entity-based & AI systems notice.

Your activity and reputation on Reddit, Quora and niche forums, where people ask and answer candidly. AI systems increasingly pull from these communities when forming answers.

Sub-topics

Niche Forums

Smaller, topic-specific communities (an industry forum, a hobby board) — very specific, candid discussion that search and AI draw on for detailed questions.

Do this

• Answer helpfully first; link only when it truly helps
• Always disclose that you work for the brand

The baseline setup to appear in local results: a complete Google Business Profile, consistent details everywhere, citations, and a page for each location you serve.

Think of it likeRelevance + distance + prominence.

Sub-topics

Google Business Profile

Free listing with hours, photos, phone & reviews in Maps and local results. The #1 local asset — claim it, fill every field.

NAP Consistency

N ame, A ddress, P hone identical everywhere. Even “St.” vs “Street” can confuse systems.

Local Citations

Mentions of your NAP on other sites & directories, even without a link — independent proof you’re real and where you are.

Local Landing Pages

One genuinely unique page per location/area: address, hours, services, local reviews, map. Not a template with only the city swapped.

Reviews & Velocity

A steady stream of new reviews looks healthy; a sudden burst looks suspicious. Ask every happy customer.

Local Links

Links from local news, associations, events and bloggers — worth more locally than random links.

Local Pack & Maps

The 3 listings + map for local searches. Driven by profile completeness, reviews & relevance.

Multi-Location SEO

Several branches or a service area? Each needs its own page and listing — not one generic entity.

BrightLocal tool

Local SEO platform: local rank tracking, citation audits and review management.

Whitespark tool

Local SEO tool for citation building and local rank tracking.

Optimising a site to serve and rank for audiences in multiple countries or languages — handling how content, URLs and targeting signals differ per market.

Think of it likeTranslate the meaning, not just the words.

Sub-topics

Hreflang

An HTML annotation telling search engines which language/region version to show each user — and that translations are alternatives, not duplicates.

ccTLD, Subfolder, Subdomain

ccTLD (example.de): strongest country signal, separate domains to build. Subfolder (example.com/de/): shares the main domain’s authority, simplest. Subdomain (de.example.com): easy to host separately, authority partly separate.

Geo-Targeting & Localization

Geo-targeting: signalling which country a page is for (hreflang, ccTLD or folder). Localization: adapting the page itself — language, currency, spelling, units, examples, payment options.

Optimising videos and their metadata so they’re found on YouTube and in Google (video results, video packs, key moments).

Think of it likeSearch engines can’t “watch” well — so describe the video in words they can read.

Do this

• Give each video its own watch page where the video is the main content
• Add VideoObject schema + a video sitemap
• Write a clear title, description & chapters (key moments)
• Publish a transcript; use a custom, readable thumbnail

Optimising an online store’s product and category pages — with retail-specific headaches like huge catalogues and endless filter URLs.

Think of it likeCategory pages are your shop windows; product pages are the shelf labels.

Do this

• Unique product descriptions (not the maker’s copy)
• Product schema + a Merchant Center feed
• Treat category pages as landing pages with helpful intro text
• Control faceted navigation (see Stage 3)
• Out of stock for good? Redirect to the closest product

Earning visibility in Google surfaces that recommend content without a typed query: Top Stories, Google News and the Discover feed.

Sub-topics

Google Discover

A personalised mobile feed based on interests, not searches. You can’t target it with keywords — quality, freshness & images decide.

Do this

• Fresh, original, well-sourced reporting
• Clear author bylines & publish/updated dates
• Large images (1200px+ wide) and max-image-preview:large
• Honest headlines — clickbait hurts in Discover

Free, and straight from Google: shows exactly how your site performs — the queries you appear for, clicks, CTR, indexing errors and penalties.

Think of it likeYour free window into Google — start every investigation here.

Sub-topics

GSC API & BigQuery Export

The API pulls data into your own tools automatically. The BigQuery bulk export streams it daily into a warehouse — more rows and history beyond the interface’s 16 months.

The core numbers for how visible you are in classic search.

Sub-topics

Impressions, Clicks & CTR

Impressions: times you were shown. Clicks: times you were chosen. Low CTR at a good position → rewrite the title/description.

Position & Share of Search

Average position: where you usually rank. Share of search: your slice of all searches in your category — the bigger picture.

Branded vs Non-Branded

“Nike running shoes” vs “best running shoes”. Shows if traffic comes from fans or from new discovery.

Query & Topic Coverage

How many relevant queries you appear for — and whether you cover the whole topic or just a corner of it.

Monitoring where you rank for target keywords over time, using tools that check results regularly.

Sub-topics

Semrush tool

Position tracking across devices and locations, with visibility trends over time.

AccuRanker tool

Fast, accurate rank tracking for large keyword sets.

Do this

• Track by device and location
• Group keywords by topic, not one by one
• Note which SERP features appear

What visitors do once they land: where they came from, which pages they view, how they interact.

Sub-topics

Google Analytics tool

Free (GA4): sources, pages, and key actions like sign-ups or purchases.

Microsoft Clarity tool

Free heatmaps & session recordings — see how people actually move through a page.

Looker Studio (Google Data Studio) tool

Now Looker Studio: free dashboards combining GA, Search Console & more.

LLM Referral Traffic

Visits from links inside AI answers (ChatGPT, Perplexity…). Track them separately to prove AI value.

Valuing searches where people get their answer on the results page or in an AI reply — without visiting you. Invisible to normal analytics, yet real brand exposure.

Think of it likeAn iceberg: clicks are the tip; impressions and AI mentions are the hidden value.

Track a fixed set of real prompts across AI systems, again and again: are you mentioned, cited, how prominently, against whom — and in what tone?

Think of it likeA prompt set is the new keyword list.

Sub-topics

Prompt Coverage & Mention

Across your prompt set, how often you appear at all — cited as a source or simply named.

Citation Rate & Position

How often you’re cited — and where. First source listed beats being buried at #6.

Competitive Answer Share

Your presence in AI answers vs competitors on the same prompts: winning, losing or level?

Entity & Sentiment

Does the AI link you to the right topics — and is the tone positive, neutral or negative?

Ahrefs Brand Radar tool

Tracks how often and how prominently a brand appears in AI-generated answers.

Profound tool

Tracks how often and how favourably a brand appears across AI systems.

Otterly tool

Monitors brand mentions and citations in AI search and chat.

Peec tool

Tracks share of voice in AI answers against competitors.

Structured experiments that check whether a change actually improved visibility — instead of assuming it did.

Think of it likeOpinions are cheap — experiments aren’t.

Sub-topics

Content & Schema A/B Tests

Change one group of similar pages, keep a matching group untouched as the control, compare. (SEO tests split pages, not users.)

Before/After Prompt Tests

Run the same AI prompts before and after a change, and compare how AI describes or cites you.

Competitor Benchmarking

Compare your search & AI visibility with competitors’ — is 20% good? Only context tells you.

A structured review of technical health, content quality and backlinks — usually the first step when a site underperforms, and the baseline before new work.