Half the marketing glossaries on the internet are stuck in 2018.
They define "impressions" the same way they did when nobody had heard of an AI Overview. They list "reach" as a vanity metric without explaining how it's calculated post-iOS 14. They explain attribution like multi-touch is something you can actually measure cleanly in 2026.
So in this guide, I'll walk you through the working glossary. 70+ terms, organised by category, with definitions that reflect how the industry actually operates now.
Acquisition + Funnel
CAC (Customer Acquisition Cost): total spend to acquire a customer, divided by customers acquired. In 2026, blended CAC (all marketing spend / all customers) matters more than channel-attributed CAC because attribution data is broken.
Blended CAC: all marketing spend (paid + content + agency fees + tools + salaries) divided by all customers acquired in the period. The honest number. Usually 30-50% higher than platform-reported CAC.
LTV (Lifetime Value): total gross profit a customer generates over their lifetime with you. "Lifetime" usually capped at 12-36 months for forecasting purposes.
LTV:CAC ratio: the leverage metric. 3:1 is healthy for most categories. Below 2:1 is unsustainable. Above 5:1 usually means under-investing in growth.
Payback period: how many months until cumulative gross profit from a customer equals the CAC paid to acquire them. Targets vary by business model: 3 months for cash-sensitive DTC, 12-18 months for VC-backed B2B SaaS.
Cohort: a group of customers acquired in the same period (usually a calendar month). Cohort analysis tracks how each group's behaviour changes over time.
Attribution
Last-click attribution: credits the final touchpoint before conversion with 100% of the value. Mostly broken in 2026 because most journeys involve multiple touchpoints across iOS-restricted environments.
Multi-touch attribution (MTA): distributes credit across multiple touchpoints. Sounds smart but data is incomplete post-iOS 14, so MTA models often produce confidently wrong numbers.
Marketing Mix Modelling (MMM): top-down statistical model that estimates contribution of each channel to revenue. Coming back into fashion as MTA degraded. Slower to react but doesn't require user-level tracking.
Incrementality testing: the gold standard. Hold out a control group, run campaigns to a treatment group, measure the lift. Slower but actually true.
Geo holdout test: incrementality test where one region runs a campaign and a comparable region doesn't. Compare conversion lift between them.
iROAS (incremental ROAS): the ROAS that's actually attributable to your ad spend, after subtracting what would have happened anyway. Usually 30-60% lower than platform-reported ROAS.
MER (Marketing Efficiency Ratio): total revenue / total marketing spend. The honest top-line number that doesn't depend on attribution working.
Paid Advertising
CPM (Cost Per Mille): cost per 1,000 impressions. The base unit of brand and reach campaigns.
CPC (Cost Per Click): cost per click. The base unit of paid search and most direct-response paid social.
CPA (Cost Per Acquisition): cost per conversion event. Same as CAC for top-of-funnel conversions; closer to true CAC at bottom of funnel.
ROAS (Return on Ad Spend): revenue / ad spend. 3:1 minimum for most DTC. iROAS is the more honest version.
CTR (Click-Through Rate): clicks / impressions. Above 1% on cold paid social is healthy; above 3% is excellent.
Conversion Rate: conversions / clicks. Varies wildly by industry and funnel stage.
Frequency: average number of times an audience saw your ad. Above 4 with declining CTR signals fatigue.
Quality Score (Google): 1-10 score Google assigns to your ad. Higher score = lower CPC for the same position.
Performance Max (PMax): Google's automated cross-channel campaign type. Combines Search, Shopping, Display, YouTube, Discovery into one campaign.
Advantage+ Shopping (Meta): Meta's automated DTC campaign type. Replaces manual targeting with algorithmic optimisation.
SEO + Content
SERP (Search Engine Results Page): the page Google shows after a query. In 2026, increasingly compressed by AI Overviews, ads, and SERP features.
AI Overview: Google's generative answer that appears at the top of many SERPs. Cites 3-5 sources. Compresses click-through rates 30-60% even when your underlying ranking holds.
Featured snippet: the boxed answer at the top of some SERPs. Less prominent in 2026 because AI Overviews replaced many.
Domain Rating (Ahrefs) / Authority Score (Semrush): third-party estimates of domain authority based on backlink profile. Useful relative comparison; not a Google metric.
Backlink: a link from another site pointing to yours. Quality matters far more than quantity.
Anchor text: the visible text of a link. "Click here" is bad anchor text; descriptive keywords are good.
Internal link: a link from one page on your site to another. Underrated SEO signal; redistributes link equity across your domain.
Crawl budget: the number of pages Google's crawler will visit on your site per day. Matters for large sites; usually irrelevant for small ones.
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness): Google's quality framework for evaluating content. Not a single ranking factor, but a strong influence on YMYL (Your Money Your Life) topics.
Helpful Content: Google's signal for content created for humans first, search engines second. Multiple updates 2022-25 demoted thin/AI-spun content.
Topical authority: the perceived expertise of your domain on a specific topic, built through depth of content coverage.
Pillar content: long-form, comprehensive content (2,500+ words) on a foundational topic, supported by cluster content.
Cluster content: shorter, more specific articles linking back to the pillar. Builds topical authority.
Email + Lifecycle
Open rate: percentage of recipients who opened your email. Inflated by Apple Mail Privacy Protection since 2021; treat as directional, not absolute.
Click rate: percentage of recipients who clicked a link in your email. The honest engagement metric.
Click-to-open rate (CTOR): clicks divided by opens. Less affected by Apple inflation; better quality signal.
Deliverability: the percentage of emails that actually reach the inbox (vs spam folder). In 2026, requires SPF, DKIM, DMARC properly configured plus BIMI for brand presence.
Welcome series: automated emails sent to new subscribers/customers. Typically the highest-revenue email flow per send.
Abandoned cart flow: automated emails to users who added to cart but didn't purchase. 10-20% recovery rate is healthy.
Browse abandonment: emails to users who viewed product but didn't add to cart. Lower-converting than abandoned cart but higher volume.
Post-purchase flow: emails after purchase. Replenishment, cross-sell, review request. Underrated as a retention lever.
RFM segmentation (Recency, Frequency, Monetary): classic customer segmentation framework. Group customers by how recently they bought, how often, and how much.
Churn rate: percentage of customers who stop buying / cancel in a period. Sustainable SaaS aims for under 5% monthly churn.
Conversion Optimisation
CRO (Conversion Rate Optimisation): the discipline of increasing the percentage of visitors who convert. Includes copy, design, UX, and offer testing.
A/B test: randomly split traffic between two variants, measure which converts better. Statistical significance requires meaningful sample sizes.
Multivariate test: tests multiple changes simultaneously. Powerful but requires very high traffic.
Heatmap: visualisation of where users click, scroll, or hover on a page. Useful for diagnosing UX problems.
Session recording: video playback of individual user sessions. Diagnostic tool for understanding user behaviour qualitatively.
Friction: anything in the user journey that reduces conversion. Form fields, page load time, unclear CTAs.
Above the fold: the part of the page visible without scrolling. Critical for first-impression conversion.
Modern AI / LLM Terms
LLM (Large Language Model): the technology behind ChatGPT, Claude, Gemini. Trained on massive text corpora to predict next token.
RAG (Retrieval-Augmented Generation): LLMs combined with a search step. Used by Perplexity and ChatGPT search to ground answers in real-time information.
llms.txt: a file at your site root that tells AI crawlers what content to prioritise. Emerging standard.
AI Overview: see SEO section above.
Prompt engineering: the discipline of crafting LLM inputs to get reliable outputs. Increasingly relevant for marketers using AI tools.
Hallucination: when an LLM generates plausible-sounding but factually wrong information. Why human review of AI output is non-negotiable.
Analytics Terms
GA4 (Google Analytics 4): replaced Universal Analytics in July 2023. Event-based, not session-based.
Engagement rate (GA4): percentage of sessions that lasted 10+ seconds, fired a conversion event, or had 2+ page views. Replaced bounce rate as the headline metric.
Bounce rate (GA4): 1 minus engagement rate. Different definition from Universal Analytics.
Conversion event: any event you've configured GA4 to count as a goal completion. Define these intentionally.
Attribution model (GA4): Data-driven attribution is the default in GA4 now, replacing last-click as the standard.
The honest summary
Marketing glossaries are useless without context. The same term means different things in different industries, and definitions evolve as platforms change.
Use this glossary as a reference, but always verify the term means what you think it means in YOUR specific context.
And remember: most marketing problems aren't solved by knowing the definitions. They're solved by knowing which metrics actually matter for your business model. Half the terms above are vanity metrics in some contexts and essential in others.
