The Ecommerce Glossary

The exact words and phrases shoppers type into Google when they want to find, compare or buy a product. The bridge between customer intent and your store.

Targeting the right keywords means your products appear in front of buyers ready to spend, not researchers. Wrong keywords = wasted content and zero revenue.

A link from another website pointing to yours. Google treats each one as a vote of confidence the more trusted the source, the more powerful the vote.

Still one of Google's top 3 ranking factors in 2026. Quality links from real publications, blogs and partner brands push product and collection pages above competitors.

The clickable blue headline that appears in Google's search results. First impression of your page in the SERP and a primary ranking signal.

A weak title kills clicks even at #1. A sharp, benefit-led title with the right keyword can lift CTR 20-60% and steal traffic from competitors above you.

The short summary shown under the title tag in the SERP. Not a direct ranking factor but the free ad copy that earns the click.

Treat it like a paid ad headline. A persuasive meta with benefit, proof point and CTA can double CTR without any ranking change.

Search Engine Results Page the page Google shows after a search. Includes ads, product listings, organic results, images, videos, AI Overviews and featured snippets.

In 2026 the organic top 3 isn't enough. You need to win across shopping, AI Overviews, reviews and rich snippets to capture real attention.

How Google's bots discover pages on your store by following links and reading your sitemap. If a page isn't crawled, it cannot rank.

Most ecommerce stores burn crawl budget on filter and tag URLs. Cleaning that up gets money pages indexed in days.

When Google stores your page in its database so it can appear in results. Crawled ≠ indexed. Indexed = eligible to rank.

Thousands of stores have hundreds of products crawled but never indexed meaning zero traffic. Fixing indexation alone unlocks instant organic growth.

The clickable, underlined text inside a link. It tells Google (and humans) what the linked page is about.

Descriptive internal anchors build topical authority and push key collections and PDPs up the rankings. "Click here" wastes link equity.

The written description of an image inside your store's HTML. Helps visually impaired users and helps Google understand what the image shows.

Wins traffic from Google Images, improves accessibility and adds keyword context to product pages without stuffing.

Structured code added to pages that helps Google understand your content products, prices, reviews, FAQs, breadcrumbs.

Unlocks rich results: star ratings, price, stock, FAQs directly in the SERP. Rich snippets lift CTR 30-80% with no ranking change.

An HTML tag telling Google which version of a URL is the master copy when duplicates exist (e.g. filter parameters, printer-friendly pages).

Consolidates ranking signals to one URL instead of splitting them across duplicates. Especially critical for Shopify collections and WooCommerce product variants.

Google's three page-experience metrics: LCP (loading), INP (interactivity) and CLS (visual stability). Measured on real Chrome users.

A confirmed ranking signal AND a conversion signal. A 1s LCP improvement can lift conversion 5-10% on mobile ecommerce.

Google's confidence that your site is a genuine expert on a topic built by depth, breadth and interlinking across related pages.

A store with 40 interconnected guides on skincare will outrank one with a single "best serum" post, even with weaker links.

Optimising pages so AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Claude, Gemini) cite you as the source of a specific answer.

AI answers increasingly replace clicks. Being the cited source captures brand, trust and referral traffic even when the click never happens.

The broader strategy of shaping how generative AI systems understand, describe and recommend your brand entity, schema, sameAs, external mentions.

AEO wins the citation on a specific answer. GEO decides whether the model recommends YOU by name for the whole category.

The percentage of visitors who complete a target action usually a purchase. Sessions with orders ÷ total sessions.

Every 0.1% CVR uplift compounds against 100% of paid and organic traffic. It's the highest-leverage number in ecommerce.

Revenue ÷ number of orders. What the average customer spends in a single checkout.

AOV improvements pass 100% to contribution margin no extra ad spend, no extra CAC. Free-shipping thresholds, bundles and upsells all target AOV.

The percentage of PDP visitors who add a product to cart. Diagnoses PDP quality independent of checkout friction.

A low ATC rate points to weak imagery, unclear pricing, missing reviews or vague delivery info problems no ad spend can fix.

The percentage of shoppers who add items to cart but leave before completing checkout.

The UK average is ~70%. Every 5% recovered abandonment is a permanent revenue line item usually cheaper than any acquisition channel.

A page purpose-built for a single audience, message and call-to-action usually the destination for a paid campaign.

Sending cold traffic to a homepage is the #1 waste of ad budget. A campaign-matched landing page can 2-4× conversion at the same CPM.

A controlled experiment that splits traffic between two versions of a page or element to see which produces more conversions.

Removes opinion and HIPPO decisions. Proper A/B testing with adequate sample size compounds into real revenue instead of guesses.

A visual overlay showing where users click, scroll or hover on a page. Common tools: Hotjar, Microsoft Clarity.

Reveals what users actually do vs what you assume. Usually finds a dead click on a non-linked hero, or a scroll cliff before the ATC button.

A replay of a real user's session clicks, scrolls, form interactions captured for later review.

The fastest way to find checkout bugs, confusing UI and rage clicks. Watching 20 recordings usually surfaces 3-5 high-impact fixes.

A small on-site action indicating buying intent short of a purchase ATC, email signup, wishlist, size-guide open, review view.

Micro-conversions give statistical power on lower-traffic stores where full-purchase A/B tests take months to reach significance.

Return on Ad Spend revenue generated ÷ ad spend. The most common ecommerce paid-media metric.

ROAS alone is misleading because it ignores product margin. A 4× ROAS on a 20% margin product still loses money after ops and shipping.

Marketing Efficiency Ratio total revenue ÷ total marketing spend across all channels. A blended, platform-agnostic view.

Platform ROAS lies (last-click, view-through inflation). MER is what your P&L actually experiences.

Customer Acquisition Cost total sales and marketing spend ÷ new customers acquired.

The other side of LTV. A profitable business needs LTV > 3× CAC on 12-month payback, or the growth is subsidised debt.

Revenue minus variable costs (COGS, payment fees, shipping, ad spend). What's left to cover overheads and profit.

The only number that tells you if a scale-up is actually earning money. ROAS and even MER can look great while CM goes negative.

Cost Per Mille the cost to reach 1,000 impressions on an ad platform. The raw auction price of attention.

Rising CPMs are the #1 reason "same creative, same audience" campaigns die. Creative refresh and audience diversification counter it.

Click-Through Rate clicks ÷ impressions. A proxy for creative and offer relevance.

Higher CTR usually means lower CPC and cheaper conversions. Creative testing is really CTR testing at scale.

The structured process of testing new ad creatives against a control to find winners that lower CPA and extend campaign lifespan.

In 2026, creative is the biggest lever in Meta and TikTok. Winning brands ship 15-40 new creative variants per week.

Lifetime Value total gross profit expected from a customer over their entire relationship with the brand.

LTV determines how much you can afford to spend acquiring a customer. High-LTV brands can outbid low-LTV competitors in every auction.

An automated, behaviour-triggered email sequence in Klaviyo welcome, abandoned cart, browse abandonment, post-purchase, winback.

Flows generate 40-60% of email revenue at a fraction of the effort of campaigns. Fixing broken flows is usually the fastest email revenue win.

The percentage of customers who buy again within a defined window (30/60/90 day).

The single biggest determinant of LTV. A 5% RPR lift compounds into significantly more affordable acquisition ceilings.

Direct promotional and transactional messages delivered to a customer's mobile via SMS or MMS, usually opted-in at checkout.

40-98% open rates and near-instant CTR beat email for time-sensitive drops, restocks and cart recovery when used sparingly.

Splitting a customer list into groups by behaviour, purchase history, geography or lifecycle stage so messages match the audience.

Blasting your full list burns deliverability and unsubscribes. Segmented campaigns typically 2-3× revenue per send.

Google Analytics 4 Google's event-based analytics platform, replacing Universal Analytics since July 2023.

The default source-of-truth for organic, referral, direct and blended attribution. Broken GA4 = broken decisions.

Sending analytics and pixel events from your own server (e.g. GTM server container) instead of the browser, then forwarding to Meta CAPI, Google, TikTok, GA4.

iOS ATT, Safari ITP and ad-blockers destroy client-side tracking. Server-side recovers 15-35% of missing conversions and stabilises pixel signal.

The rules that decide which touchpoint gets credit for a conversion last-click, first-click, linear, time-decay, data-driven.

Every channel looks like a hero under last-click and a villain under first-click. Choosing (and standardising on) one model is more important than which one.

Tags appended to URLs (utm_source, utm_medium, utm_campaign) that let analytics tools identify where a click came from.

Without disciplined UTMs, email, affiliate and paid social traffic gets misattributed as "direct" or "referral" hiding what actually works.

Shopify's enterprise tier for brands scaling past ~£1M/year revenue that need Launchpad, Scripts (now Functions), B2B, custom checkout, multi-store.

The extra features matter only above a certain scale. Below £1M ARR most brands pay for capability they never use.

An architecture where the storefront (frontend) is decoupled from the ecommerce engine (backend) via API e.g. Shopify Hydrogen, Next.js + Storefront API.

Unlocks best-in-class performance, custom UX and multi-channel commerce. But it adds engineering cost most brands cannot justify.

Shopify's modern (post-checkout.liquid) way to customise checkout via checkout UI extensions, apps and Shopify Functions.

Old checkout.liquid customisations are being deprecated. Migrating is not optional if you want upsells, custom shipping logic or delivery date pickers at checkout.

The open-source ecommerce plugin for WordPress flexible, low-cost, but self-hosted and dependent on plugin discipline for performance and security.

Perfect for content-led, high-margin niches that want full control. Painful for brands scaling without engineering headcount.

Product Detail Page the individual product page where a shopper decides whether to buy.

The highest-value page in ecommerce. A 5% PDP CVR uplift compounds across every acquisition channel that lands there.

Product Listing Page the collection or category page listing multiple products.

PLPs are the second-highest-value template. Fast, well-filtered PLPs let shoppers self-serve and lift discovery-driven conversion.

Google's generative answer box that appears above the traditional 10 blue links for many searches, powered by Gemini.

Reduces clicks to organic listings for informational queries. The winning strategy is to be cited inside the Overview, not fight it.

When ChatGPT, Claude, Gemini, Perplexity or Copilot references your page or brand as a source when answering a user's question.

Citations = trust + qualified referral traffic + brand memory in the model itself. Increasingly the top-of-funnel channel in 2026.

A plain-text file at your domain root that gives large language models a curated map of the most citation-worthy pages on your site.

It doesn't guarantee crawling but signals editorial priority helping models pick the right canonical URL when they cite you.

The discipline of designing inputs to LLMs instructions, context, examples, constraints so the model produces reliable, on-brand output.

The difference between AI content that ranks and AI content that gets penalised is usually the prompt, not the model.

Optimising for how search engines and LLMs model your brand as an entity in a knowledge graph with a canonical name, description, sameAs profiles and relationships to other entities.

Google's Knowledge Graph and LLM training data both reward disambiguated, well-connected entities. Weak entities get confused with competitors.

Generating hundreds or thousands of high-quality pages from a structured data source and a template used for locations, integrations, comparisons.

Captures long-tail search demand no manual editor could economically cover. Only works when template quality is genuinely useful, not spun.

Search that matches meaning rather than exact keywords by comparing vector embeddings of the query and documents.

Powers modern site search, product discovery and internal knowledge bases surfacing results that keyword search misses.

Optimising a website so search engines and AI systems recognise it as a distinct, well-defined entity brand, founder, service rather than a bag of keywords.

AI Overviews and LLMs cite entities they can identify with confidence. A strong entity graph turns your brand into a nameable source, not a URL.

The perceived expertise a domain has around a subject, built through comprehensive, interlinked content covering every angle of that topic.

Google rewards sites that are the definitive source for a topic. Ten interlinked ecommerce SEO guides outrank ten scattered posts on unrelated subjects.

Structuring content so answer engines (Google AI Overviews, ChatGPT, Perplexity, Claude) can extract, cite and quote it directly.

In 2026 an increasing share of queries never reach a website. Being the source AI answers cite is the new position #1.

The discipline of getting your brand named, quoted and linked inside LLM responses through verifiable stats, named authors and outbound citations.

LLMs prefer content with clear provenance. Named experts and cited sources are the currency of the generative web.

Experience, Expertise, Authoritativeness and Trustworthiness the framework Google uses to assess whether content is worth surfacing, especially for YMYL and expert topics.

Every commercial ecommerce query is now evaluated on E-E-A-T. Weak founder bios, no authors, no reviews = capped rankings.

Google Search Generative Experience the AI-generated answer that appears above the classic ten blue links for many queries.

AI Overviews can absorb 30-70% of clicks. Ranking inside them, not just below them, is the modern SEO priority.

The single URL you tell Google to treat as the master version of a page when the same content is reachable through several URLs.

Ecommerce faceted navigation, tracking params and pagination create dozens of duplicate URLs per PDP. Canonicals stop the duplicate content from splitting ranking.

Google's three page-experience metrics LCP (largest contentful paint), INP (interaction) and CLS (layout shift) measured on real user visits.

Fail Core Web Vitals and rankings quietly slide. Fix them and conversion rate typically lifts 5-12% alongside the ranking gain.

Reading a site's server logs to see exactly which URLs Googlebot crawls, how often, and how much crawl budget is wasted on parameter or filter URLs.

The only source of truth for crawl behaviour. Reveals index bloat, wasted crawl budget, and orphan PDPs that GSC and crawlers cannot show.

Filters (size, colour, price, brand) that generate URL variants a source of crawl waste and duplicate content when unmanaged.

Every ecommerce SEO scale problem eventually traces back to faceted navigation. Handling it correctly is the difference between 200 and 20,000 indexed URLs.

Generating a large number of high-quality pages from structured data and a template used for locations, integrations and comparison sets.

Captures long-tail search demand no editorial team could economically cover. Only works when the template is genuinely useful.

HTML tags that tell Google which language and country version of a page to serve the core of international SEO.

Wrong hreflang serves the US site to UK shoppers and tanks conversion. Correct hreflang keeps each market's SEO cleanly separated.

The bordered answer box at the top of the Google SERP that quotes a short passage, list or table from a ranking page.

Owning the snippet can lift organic CTR by 20-30% even without rank change. AI Overviews increasingly source their answer from the featured snippet.

The way pages on your own site link to each other the strongest ranking signal you fully control.

Ecommerce sites with disciplined hub → guide → PDP internal linking outrank competitors with more backlinks. It's the cheapest SEO lever.

A permanent server-side redirect from an old URL to a new one, passing nearly all ranking signal to the destination.

Every replatform, PDP rename or category restructure without a full 301 map costs organic revenue. Migration success = redirect map quality.

A plain-text file at the site root that tells crawlers which URLs they may or may not fetch the crawl-budget kill switch.

One wrong line disallows Google from your whole store overnight. One right line frees crawl budget for the URLs that matter.

Enhanced search results with review stars, price, availability, FAQs or breadcrumbs powered by schema markup.

Rich results steal clicks from plain listings above them. Product schema alone can lift ecommerce SERP CTR by 20-35%.

The confidence level that an A/B test result is real rather than noise usually a 95% or higher probability the winner beat the control.

Calling wins too early is the most common CRO mistake. Un-significant results shipped as "wins" waste months of optimisation budget.

The number of visitors or sessions needed for an A/B test to detect a meaningful lift with confidence.

Testing on the wrong sample size guarantees false wins or missed real wins. Sample size must be planned before the test starts.

A visual overlay showing where visitors click, scroll or hover on a page used to spot friction, dead zones and misread elements.

Reveals why a page underperforms when analytics only says it does. Direct input for CRO hypotheses.

Recorded playback of a real visitor's session every mouse move, click and scroll used to diagnose friction and bugs.

Faster than any survey. Ten replays of cart abandoners typically reveal three fixable friction points.

A smaller conversion event on the path to a purchase add to cart, wishlist add, email signup, video watch.

Traffic rarely converts on first visit. Optimising micro-conversions lets you improve the funnel without waiting for enough purchases.

The one-sentence answer to "why buy this from you, right now, at this price?" the highest-leverage element on any commercial page.

Fixing a weak value prop lifts conversion more than any button colour test. It's also the cheapest change.

The part of a page visible without scrolling on a first view the region that decides whether a visitor stays or bounces.

60% of visitors never scroll past the fold. Everything critical headline, product hero, price, primary CTA has to live there.

Total revenue ÷ total ad spend across all channels. The single blended metric that platform ROAS reporting cannot manipulate.

Platform ROAS double-counts. MER doesn't. Every mature DTC brand eventually reports on MER as their scaling scorecard.

Revenue minus variable costs (COGS, fulfilment, ad spend, payment fees) the profit each order generates before overhead.

ROAS ignores gross margin; contribution margin is the number your CFO cares about. It's the honest ceiling on paid scaling.

The true cost of acquiring a new customer total marketing spend divided by new customers, blended across all channels.

Platform CAC lies. Blended CAC over LTV is the ratio that decides whether growth is profitable at scale.

Google's automated campaign type that serves ads across Search, Shopping, YouTube, Discover, Display and Gmail from a single asset group.

70%+ of Google Shopping spend now runs through P-Max. Understanding asset groups and audience signals is basic paid-media literacy.

Server-side event API for Meta (and equivalents on Google, TikTok, Snap) that sends conversion data directly, bypassing iOS and browser signal loss.

iOS 14+ killed 20-30% of pixel signal. CAPI restores match rate and lets Advantage+ Shopping actually optimise.

Meta's fully automated shopping campaign product one campaign, up to 150 creatives, algorithmic audience and placement selection.

For most DTC brands under £250k/month ad spend, Advantage+ outperforms manual structures. Fighting it is expensive.

The disciplined weekly rhythm of shipping new ad creatives usually 15-40 variants per week to keep learning phase filled and CPMs contained.

Meta and TikTok reward volume of creative variance more than any single "perfect" ad. Brands testing weekly outperform brands testing monthly by ~40%.

The extra revenue paid media actually generates above the baseline that would have happened without it measured through geo-holdouts or MMM.

Platform ROAS overstates paid contribution 30-70% by claiming credit for organic and repeat revenue. Incrementality is the correction.

A statistical model that attributes revenue across channels using time-series data the modern replacement for last-click attribution.

As deterministic tracking degrades, MMM is the only defensible way to plan channel spend at £250k+/month.

Segmenting customers by Recency, Frequency and Monetary value the classic model behind every mature ecommerce email programme.

RFM identifies the top 20% of customers driving 60%+ of revenue. Treating them differently is the highest-leverage retention move.

Model-predicted lifetime value for each customer, based on early purchase behaviour used for bid ceilings and VIP flows.

Averaging LTV across all customers wastes budget. Bidding to high-predicted-CLV customers unlocks scaling headroom rivals cannot see.

The share of your emails that actually reach the inbox rather than the promotions tab or spam folder.

Every 1% drop in inbox rate costs roughly 1% of email revenue. Below 90% inbox, your programme quietly bleeds.

An automated segment that unsubscribes disengaged subscribers after a defined window the deliverability hygiene lever.

Emailing dead subscribers destroys sender reputation and drags every send down. Sunsetting looks like list shrinkage but protects revenue.

Segmented automations targeting customers who haven't bought in 30/60/90 days the highest-leverage retention channel after post-purchase.

Reactivating a lapsed customer costs 5-7× less than acquiring a new one. Well-built winback flows drive 3-6% of email revenue.

GA4's structured event model for ecommerce view_item, add_to_cart, begin_checkout, purchase that unlocks funnel reporting and audience building.

Without enhanced ecommerce events, GA4 is blind to your funnel. Every dashboard, audience and attribution decision depends on these events firing correctly.

Sending tracking data from your server (via GTM Server, Stape or similar) instead of directly from the browser restoring signal lost to ad blockers and iOS.

Recovers 15-35% of conversion signal lost client-side. Pays for itself in improved bid quality within weeks at £20k+/month ad spend.

The lookback period a platform credits itself with the conversion commonly 7-day click / 1-day view on Meta, 30-day click on Google.

Different windows report wildly different ROAS for the same campaign. Understanding the window is the first defence against being fooled by dashboards.

Google's framework for adjusting measurement based on user consent required in the EEA/UK for advertisers to keep receiving conversion signals.

Non-compliant sites lose access to Google Ads audiences and Enhanced Conversions. Compliant sites keep bid quality and remarketing intact.

Shopify's enterprise tier checkout extensibility, up to 200 markets, B2B, launchpad, scripts and dedicated account support.

Above £1-2M/year, Plus features (checkout blocks, B2B, multi-market) typically pay for themselves. Below that, Advanced is usually enough.

Decoupling the storefront (front-end) from the commerce engine (back-end) usually via Hydrogen (Shopify) or a custom Next.js build on Woo/BigCommerce.

Unlocks best-in-class performance and freedom of design, at the cost of dev headcount. Rarely worth it below £5M/year.

A hosted ecommerce platform competing with Shopify stronger native B2B and multi-storefront, weaker app ecosystem.

Best fit for wholesale-heavy or complex catalog brands that outgrow Woo but don't want Shopify's rented model.

An architecture pattern where storefront, CMS, search, cart and checkout are best-of-breed services stitched together via APIs.

Maximum flexibility for enterprise brands with real engineering. For most SMB DTC, it's overkill and slower to ship.

An AI pattern that fetches relevant documents from your own knowledge base and feeds them to the LLM before it answers grounding responses in your data.

Stops hallucinations for internal tools (support, ops, reporting). Turns an LLM into a company-specific expert.

The craft of writing instructions to LLMs role, context, examples, constraints to get reliable, high-quality output.

Same model, same task, wildly different quality depending on prompt design. Prompt discipline is a core operator skill in 2026.

A conversational AI interface usually powered by an LLM plus a knowledge base that handles product questions, sizing, order status and returns.

Deflects 30-60% of Tier-1 tickets while lifting PDP conversion by answering pre-purchase questions live.

Training a base LLM on your own domain data to shift its default style, tone and knowledge toward your brand.

Rarely worth it in 2026 RAG plus strong prompts covers 95% of ecommerce use cases at a fraction of the cost.