From Performance Max and the Conversions API to Wikidata and the Knowledge Graph. Every definition comes from our daily practice, with benchmarks from real UA projects and links to canonical sources.
Google's main paid advertising platform: Search, Shopping, Display, YouTube, Discover, App campaigns. Known as AdWords until 2018. In UA e-commerce it is the primary channel, taking 60-80% of the paid budget. UPLIFY manages 44+ active Google Ads projects.
Product ads with photo, price, and store name in search results and on the Shopping tab. Runs on the Merchant Center feed. Since 2022 gradually replaced by PMax; in 2026 standalone Shopping is used for brand-pure campaigns and specific assortment groups.
Paid advertising across Meta platforms: Facebook, Instagram, Messenger, WhatsApp, Audience Network. Targeting via demographic, interest, and behavioral segmentation. Pixel + Conversions API is the standard tracking stack. The second-largest UA e-commerce channel by budget after Google Ads.
A Google Ads campaign type where the algorithm automatically optimizes targeting, creative, placements, and bids toward a single goal. Combines Search, Shopping, Display, YouTube, Discover, and Gmail in one "black box" campaign. Launched in 2021. Best result on UA e-commerce in 2026: median ROAS of 9.88x according to our data.
Telegram's official ad platform: short text ads with a CTA button in channels with 1000+ subscribers. No cookies — targeting is based on channel categories. Since 2024 the entry threshold has dropped to EUR 2 CPM. A niche but highly engaged channel for Ukrainian businesses.
Advertising on TikTok: In-Feed Ads, Spark Ads (boosting organic posts), catalog-based Video Shopping Ads, Smart Performance Campaigns. Tracking via TikTok Pixel + Events API. TikTok Shop is unavailable in Ukraine as of 2026 — ads drive traffic to the store website.
Advantage+ Shopping is Meta’s automated e-commerce campaign type that optimizes audiences, placements, and delivery.
Bot Ads send users to a Telegram bot — useful for lead generation, registrations, consultations, or support.
Broad audience is a wide audience without strict targeting — the algorithm finds buyers itself.
Cart abandoners are users who added a product to the cart but did not complete the purchase.
Catalog Ads use a product catalog to automatically promote items without creating a separate video for each product.
Creative fatigue happens when an audience sees the same creative too often, causing weaker performance over time.
CRM Match allows targeting or excluding users based on a customer list in a privacy-safe way.
DPA (Dynamic Product Ads) automatically shows users products from the catalog they’re most likely to buy.
Intent-based traffic comes from users who are already searching for a topic, service, brand, or competitor.
Learning phase — the period when the algorithm gathers initial conversions and stabilizes (typically 7–14 days).
Lookalike is an audience similar to your existing customers or website visitors. Meta finds users with similar behaviour.
Prospecting means advertising to new cold audiences that have not interacted with the brand yet.
Retargeting means showing ads to people who already visited the website, viewed products, added to cart, or interacted with the brand.
Search Ads appear in Telegram search results based on keywords, topics, or intent.
Whitelist and blacklist are lists of channels where ads should either be shown or excluded.
The rule for distributing conversion value across touchpoints: Last Click, First Click, Linear, Time Decay, Position-Based, Data-Driven (ML). In GA4 + Google Ads, Data-Driven has been the default since 2023.
Serverless data warehouse from Google Cloud. Standard GA4 integration — daily export of raw events. SQL analysis beyond the GA4 UI: arbitrary cohorts, custom funnels, joins with CRM/Ads data. Often indispensable for stores with $50k+/month revenue.
Server-to-server event delivery from your backend to Meta — bypasses iOS 14+ ATT and blockers. The best configuration is Pixel + CAPI with event_id deduplication. Without CAPI, Meta loses 30-50% of iOS signals. Analogs: Events API (TikTok), Enhanced Conversions (Google).
The current version of Google Analytics (since 2020), replacing Universal Analytics. Event-based model instead of session-based: every action is an event with parameters. Cross-platform (web + app), ML predictions, integration with Google Ads and BigQuery. The analytics standard for UA e-commerce.
A mechanism for conditional tag firing based on cookie consent. Default-denied sends no identified data but allows modeled conversions via ML aggregation. On uplify.agency: analytics_storage defaults to denied and switches to granted after a banner click.
A system for managing website tags without editing code: GA4 events, Meta Pixel, Google Ads conversions, TikTok Pixel, custom HTML/JS — all through a UI. Critical for cookie consent: tags do not fire until the user gives consent.
A JavaScript snippet that sends user-side events (ViewContent, AddToCart, Purchase) to Meta Ads Manager for optimization and retargeting. Blocked by iOS ATT and ad blockers — without CAPI it covers only ~50-60% of iOS conversions. Installed via GTM.
CAPI dropout happens when server-side events stop sending or become unstable.
CAPI-ready means a website or system is prepared to send server-side events to Meta through Conversion API.
Cross-channel attribution analyzes how different channels (TikTok, Meta, Google, organic) contribute to a single purchase.
Cross-channel RCA analyzes how changes in one channel (e.g. Google Ads) may affect another channel or the overall business result.
Dead clicks are clicks on elements that do nothing and may indicate UX or design issues.
Dedup rate shows whether Meta correctly merges the same event from Pixel and CAPI instead of counting it twice.
Device / Placement breakdown splits results by device and placement (Instagram Reels, Facebook Feed, Stories, etc.) so you can see what actually performs.
Drill-down is the act of zooming from a high-level view down to a specific campaign, ad group, product, or ad to see what really moved.
E-commerce events are store actions such as product view, add to cart, checkout, purchase, and order value.
EMQ-score is Meta’s event match quality score. A higher score helps Meta match conversions to users more accurately.
Enhanced Conversions improve Google Ads tracking by sending more accurate conversion data in a privacy-safe way.
Events API sends conversion events from the server to TikTok and improves tracking accuracy.
Heatmaps show where users click, scroll, and pay attention on a page.
Microsoft Clarity shows user behaviour through session recordings, heatmaps, rage clicks, and other UX signals.
Pixel / CAPI parity checks whether browser events and server events are both being sent correctly and consistently.
Quick backs are fast returns to search after clicking — often a sign the page didn’t match intent or the UX is weak.
Rage clicks are repeated clicks in one place, often showing that a user expected something to work but it didn’t.
RCA means root cause analysis. It explains not just what dropped, but why it happened and what should be checked next.
Session recordings show how users move through a website and where they face friction.
TikTok Pixel is a website code that sends TikTok data about user actions: product views, add to cart, checkout, and purchase.
Transaction ID is a unique order number that helps avoid duplicate purchase tracking in analytics.
Total revenue divided by the number of orders. The basis for break-even ROAS, target CPA, and traffic profitability estimates. Typical UA e-commerce AOV in 2026: fashion 800-1500 UAH, electronics 3000-15000, furniture 5000-30000, kids' goods 600-1200.
Spend divided by the number of conversions. A conversion is a sale, lead, or another target action, depending on setup. UPLIFY benchmark (PMax UA e-commerce 2026): median CPA 101.46 UAH. Purchase-only CPA is usually higher — 200-400+ UAH.
Clicks divided by impressions x 100%. Shows how relevant the ad is to the audience. UPLIFY PMax e-commerce benchmark: median 2.25%. Search campaigns are typically higher (5-10%) thanks to intent traffic. CTR below 1% signals a mismatch between creative and audience.
Average revenue from one customer over the entire relationship. Lets you pay more for acquisition than the first purchase brings — if the customer returns. UPLIFY guideline: a healthy LTV/CAC is 3:1 or higher.
Ad revenue divided by ad spend. Expressed as a multiplier (5x) or a percentage (500%). The key performance metric for e-commerce. UPLIFY benchmark (41 UA projects, 2026): median PMax ROAS 9.88x. To be profitable, ROAS must exceed break-even (100% / margin%).
Conversion value is the value of a conversion (e.g. order amount). It lets you optimize for revenue, not just for purchases.
DRR is the share of ad spend in revenue. A DRR of 10% means that 10 UAH out of every 100 UAH of revenue was spent on advertising.
Hook Rate is the share of viewers a creative “hooks” in the first seconds — typically how many keep watching past 3 seconds.
UPLIFY Share measures the contribution of UPLIFY-managed channels to overall business results vs organic and other sources.
VTR (View-Through Rate) shows the share of viewers who actually watched the video — a quick read on whether the creative holds attention.
The basic unit of a PMax campaign: a set of creative (headlines, descriptions, images, videos, logos) + audience signals + URL expansion. Up to 100 groups per campaign, each for a separate product/audience combination. The analog of an Ad Group in classic Google Ads.
Hints for the algorithm about whom to target in the first days of learning: Custom Segments, Customer Match (CRM lists), interests, demographics. Without signals, learning takes longer (7-10 days vs 3-5). Per UPLIFY data, the right signals add +30-50% ROAS in the first 30 days.
A list of brands PMax must not show for. Without it, PMax "eats" branded searches (your own and competitors') that would convert without ads anyway — inflating blended ROAS. UPLIFY practice: an exclusion list of 50+ brands for UA e-commerce.
A family of strategies where the algorithm sets the bid in real time toward a goal: tCPA, tROAS, Max Conversions, Max Conversion Value. PMax supports only Smart Bidding. For e-commerce the standard is tROAS.
Asset completeness checks whether a campaign has enough headlines, descriptions, images, videos, and other assets for proper optimization.
Catch-all detection finds campaigns or asset groups that take too much budget without clear segmentation and may reduce efficiency.
The product feed management platform: attributes (id, title, price, availability, GTIN, brand...) — the foundation of Shopping and PMax. Goes through Google moderation: a disapproved feed = no ads. Has official integrations with Horoshop, Shopify, WooCommerce.
A Ukrainian SaaS platform for online stores (since 2014). Built-in integrations with GMC, Meta Catalog, Google Ads, Hotline, Prom. UPLIFY is a Horoshop Expert Partner: we build and launch stores ready for advertising from day one.
A structured file (XML / CSV / Sheets / JSON) with product data — the foundation of Google Shopping, PMax, Meta Catalog, TikTok VSA. Required attributes: id, title, description, image_link, price, availability, link, brand. Feed quality is the strongest predictor of PMax ROAS.
Attention cards highlight important changes, risks, or issues so the client and manager don’t have to search for them manually.
Audit log is a history of system actions: who changed what, when, and why — needed for transparency and security.
Business Manager is the Meta hub for managing ad accounts, pages, pixels, domains, catalogs, and permissions.
Catalog Manager is Meta’s tool for managing product catalogs used in dynamic ads.
The client dashboard shows campaign performance, analytics, Merchant Center status, reports, and key business metrics in one place.
Feed rules are processing rules that automatically transform or enrich product feed data before it reaches Merchant Center.
GTIN is a global product identifier that helps Google recognize products and match them correctly in Shopping ads.
Health-check is a regular review of account health, tracking, feeds, integrations, tokens, and key project systems.
Merchant Health shows the condition of your product catalog: approved products, disapprovals, warnings, and technical issues.
QA checklist is a pre-launch list of checks for pages, forms, analytics, payments, delivery, SEO, and technical settings.
SKU is an internal product code used to manage assortment, stock, feeds, and analytics.
TikTok Catalog Manager is the tool for uploading and managing the product catalog used in TikTok Ads.
UPLIFY OS is our internal agency operating system that brings together ad accounts, analytics, alerts, reports, finances, and AI recommendations.
Webhooks automatically send data between systems when an event happens, such as a purchase or status change.
XML / CSV processing means preparing product files for import into a store or advertising platform.
Optimizing content for citation by AI engines: ChatGPT, Claude, Perplexity, Google AI Overviews, Gemini. Unlike SEO (SERP positions), the focus is entity recognition, schema readability, and citation-friendliness. A separate UPLIFY service.
An AI-generated answer in Google Search that combines citations from several sites. Launched in 2024. Appears above classic results and absorbs 30-60% of traffic. Getting cited in AIO is a key GEO goal.
A JSON-based structured data format — Google's recommended way to emit Schema.org. Embedded via script type="application/ld+json" in the head or at the end of the body. On uplify.agency every page carries 4-6 JSON-LD blocks.
Google's base of structured knowledge about billions of entities (since 2012). Integrated with Search, AIO, Maps, Assistant. Wikidata is one of its data sources: after submission, entity ingestion into the KG typically takes 2-8 weeks.
AI search with source citations in every answer (since 2022). One of the fastest GEO channels: per UPLIFY's benchmark, structured entity-recognized content appears in results within 1-3 weeks after indexing.
A shared structured data vocabulary (Google, Microsoft, Yahoo, Yandex, 2011): types like Organization, Product, FAQPage, Article with properties. Emitted via JSON-LD. The foundation of rich snippets, AI citations, and entity recognition.
A Wikimedia project of structured open data about entities. Every entity has a Q-ID. The canonical sameAs source for AI engines and Google Knowledge Graph. Submitting an organization to Wikidata is a key entity-recognition step in GEO.
AI visibility shows how easily your brand can be found, understood, and recommended by AI search systems.
Article / BlogPosting schema marks up articles, guides, and blog posts.
Brand mentions are references to your brand on other sites, in media, directories, ratings, or partner materials.
An entity is something search and AI systems can recognize: a company, person, service, product, or brand.
An entity profile is a clear representation of your brand: who you are, what you do, who you help, and why you can be trusted.
FAQPage schema marks up questions and answers so search engines can better understand the FAQ section.
LLM-citation means an AI model uses your site or brand as a source in its answer.
Organization schema describes a company: name, website, logo, contacts, social profiles, and other key details.
Person schema describes an expert or team member: name, role, image, profiles, experience, and topics of expertise.
Product + AggregateRating marks up a product with rating, price, availability, or reviews.
Proof blocks are trust signals: cases, reviews, certificates, photos, numbers, processes, and other evidence of real expertise.
Quotable claims are short, accurate, evidence-based statements that AI systems can easily quote or summarize.
Selection frameworks are structured criteria that help users and AI systems compare and choose between solutions.
Semantic coverage shows how fully a site covers a topic: services, questions, cases, comparisons, proof, and related intents.
Semantic expansion grows the topics and queries your site can be found or quoted for.
Service schema helps search engines understand what service the company offers and who it is for.
Source attribution ties a fact to its source: a date, a case, a page, an author, or another piece of evidence.
Speed metrics that affect ranking: LCP < 2.5s, INP < 200ms (replaced FID in 2024), CLS < 0.1. Measured on real users via CrUX. Poor CWV means lower positions and more expensive traffic.
Google's framework for assessing content quality: Experience, Expertise, Authoritativeness, Trustworthiness. Extended with "Experience" in 2022. Critical for YMYL topics. UPLIFY ensures it via named author + reviewedBy, Wikidata anchors, and a publishing methodology.
Breadcrumbs show the user’s path on a website, such as Home → Category → Product.
Content Gap means topics or search intents competitors already cover but your site doesn’t address yet.
Domain Authority is an aggregate score of a domain’s strength based on backlinks, mentions, and overall trust signals.
DR / UR are Ahrefs metrics — Domain Rating and URL Rating — that score the strength of a domain and an individual page.
Google Search Console helps monitor indexing, search queries, clicks, positions, and technical search issues.
Keyword-density trap is the mistake of stuffing pages with keywords instead of providing useful facts and proof.
Link profile is the totality of external links pointing to your site — their count, quality, relevance, and momentum.
Meta-tag verification is a way to confirm site ownership through a special code embedded in HTML.
Sitemap.xml is a file that lists important website pages and helps search engines discover them faster.
SSL protects website connections through HTTPS and is required for user security and stable advertising.
OpenAI's conversational AI assistant (since November 2022). Web Search was added in 2024 — it began citing sources. A key GEO target: per UPLIFY's benchmark, the gap between publication and first citation in browse mode is typically 2-8 weeks.
An AI assistant by Anthropic (since 2023): the Claude 3 → 4 → Opus 4.x series (2026). Web Search added in 2025. One of the key GEO targets for tech/marketing topics.
AI Action Plan is a draft list of recommended next steps based on campaign, analytics, product, and tracking data.
AI agents are internal assistants that check data, detect issues, prepare draft insights, and help managers respond faster.
AI analytics uses data from ads, website analytics, Merchant Center, and other sources to detect problems and growth opportunities faster.
AI Project Intelligence is an AI layer that analyzes the project across ads, products, channels, analytics, and possible growth opportunities.
Asset Quality Evaluator checks ad assets for weak copy, missing formats, duplicates, poor coverage, and relevance to the product or offer.
Client Update Draft is an AI-assisted draft explanation for the client: what happened, what it means, and what the next steps are.
Merchant fix-drafts are AI-assisted suggestions for fixing product issues in Google Merchant Center: titles, attributes, GTINs, brands, or policy warnings.
Meta QA evaluator automatically checks Pixel/CAPI, events, funnel structure, placements, audiences, and budget risks.
PMax QA is an automated check of Performance Max campaigns: tracking, assets, audience signals, structure, and budget risks.
Tools nearby: calculate ROAS and the minimum PMax budget on these same benchmarks.