Reference

Glossary.

Operator vocabulary used across the site. Plain-English definitions for terms in playbooks, frameworks, and engagements — each linked to where it lives in context. Filter by topic, or deep-link any term.

GTM + commercial

ACV: Annual Contract Value. The total annualized value of a single customer contract.
CAC: Customer Acquisition Cost. Fully-loaded spend (sales + marketing + tools) divided by new customers acquired in a period.
CLTV / LTV: Customer Lifetime Value. Expected gross-margin revenue from a single customer across their full lifecycle.
Growth Operating System: The operating model that connects revenue targets, market signals, buyer narrative, pipeline creation, sales feedback, customer proof, and growth cadence into one repeatable system.
ICP: Ideal Customer Profile. The segment where conversion, ACV, sales cycle, and retention are all best.
MSales: A Marketing-Sales operating construct that aligns marketing and sales on the same incentive plane, revenue target, buyer signal system, proof loop, and operating cadence. Also called SMarketing, M-Sales, Sales-Marketing. Used in: Performance & Native Retention & Expansion
NDR: Net Dollar Retention. (Starting ARR + expansion − contraction − churn) / starting ARR. Post-sale operating outcome. Used in: Growth Operating System
NRR: Net Revenue Retention — typically used interchangeably with NDR.
Retention & Expansion: The operating model for reducing churn, improving renewal confidence, identifying expansion triggers, and turning post-sale signals into account growth. NDR is the board-level metric it improves.
SLA (in revenue): Service-Level Agreement between marketing and sales: speed-to-lead, accepted-lead rate, first-meeting conversion, reason codes.
SMarketing: A Sales-Marketing alignment model where marketing and sales operate as one revenue system rather than separate handoff functions.

Org + operating model

Capacity model: Math tying headcount to revenue, onboarding, renewal, and product load. Refreshed quarterly.
Decision rights: Documented matrix of who has final call on recurring decision types. Replaces founder-as-router.
Global / local matrix: Three-lane model defining what global owns, what local owns, and what is shared across regional teams.
Operating cadence: The rhythm of decisions: weekly forecast, monthly NDR, quarterly board prep. Designed around decision types, not status updates.
RACI: Responsible · Accountable · Consulted · Informed. Used to make customer-journey or commercial handoffs explicit.
Role scorecard: 90/180/365-day outcome targets tied to a role. Replaces generic job descriptions.

Data + clean rooms

1PD richness: First-party data richness. The combined assessment of breadth, depth, governance, and outcome-linkage of the data the buyer controls.
ADH: Google Ads Data Hub. Privacy-safe Google / YouTube measurement environment — queries run on Google event-level data joined with first-party data, but results must clear privacy checks (default noise injection ~20 unique users per row) and cannot export user-level rows. Not a generic clean room replacement.
Clean room: Governed environment for matching, measuring, modeling, or activating data between two or more parties without exposing raw data. Used in: Data Clean Rooms, PETs & PAIR AWS Clean Rooms InfoSum
Clean room workflow: The end-to-end pattern inside a clean room — data in, match logic, query, aggregate, approval, output policy, export. The workflow, not the platform, is what ships value.
Collaboration canvas: Twelve-field artifact: ambition, business benefit, customer benefit, data in, legal basis, match logic, data out, KPI, production path, semantic context, evaluation loop, agent/API exposure.
Data collaboration stack: A seven-layer model spanning signal, governance, discovery, collaboration, metric and semantic, decision, and agentic layers. Vendors that operate across all seven win infrastructure.
Data discovery: The catalog, metadata, ownership, business domains, and definitions layer that tells teams what data exists, what it means, who owns it, and which decisions it can improve.
DCR: Data Clean Room. Generic term covering Snowflake, Databricks, AWS Clean Rooms, LiveRamp Clean Room, InfoSum, and Google ADH. IAB Tech Lab's Data Clean Room Guidance v1.0 (July 2024) defines common DCR principles, functions, and guardrails — PAIR and ADMaP are the companion interoperability protocols, not clean rooms themselves. Used in: Data Clean Rooms, PETs & PAIR
Governed output: Approved aggregates, scores, or activations that can leave a collaboration environment under explicit policy. Distinct from secure sharing and from open access.
Lineage: Auditable record of how data moved from source through transformations, governance decisions, queries, and outputs. Required for clean room and agent-ready governance.
Output policy: Approved rules defining what can leave a collaboration environment (query, aggregate, export, activate, retain) and under what thresholds. Used in: Data Clean Rooms, PETs & PAIR Multi-cloud orchestration Amazon Marketing Cloud
PAIR: Publisher Advertiser Identity Reconciliation. Originally developed by Google and donated to IAB Tech Lab, now an open standard (v1.1, finalized July 2025) for matching and activating an advertiser's and a publisher's first-party audiences without third-party cookies — join keys are commutatively encrypted and matched without decryption, then activated via encrypted IDs in OpenRTB. Supersedes OPJA; a matching-and-activation protocol, not a general analytics clean room. Used in: Google / GMP / ADH Prebid & Header Bidding
Secure sharing: Data shared with named partners under contract — but downstream use is largely unmonitored. One step beyond open access; weaker than a clean room or governed output.
Walled-garden measurement: Platform-owned, ecosystem-bounded measurement environments (Google ADH, Amazon Marketing Cloud, Meta Advanced Analytics). Distinct from open / neutral collaboration.

Semantic + metric layer

Business metric definition: Shared, governed definition of a metric (e.g. revenue, conversion, LTV, exposure) including source logic, owner, refresh cadence, and approved calculation. Prevents the same word meaning three things across teams.
Metric governance: The set of rules and owners that maintain shared metric definitions across teams. Without it, dashboards, planning, and agentic answers will not agree with each other.
Semantic layer: The curated layer between raw tables and business users — metadata, definitions, synonyms, value mappings, approved questions, example SQL. Where data becomes usable in business language. Used in: Semantic Infrastructure Databricks
Value dictionary: A mapping from natural-language terms a business user might use to the actual fields and values they correspond to. The bridge between "show me top customers" and the right SQL.
Embedding: A learned vector (list of numbers) standing in for an object — text, image, audio, a user or content profile — so distance between vectors approximates similarity of meaning.
Vector: An ordered list of numbers; the numeric form an embedding takes.
Vector space: The multi-dimensional space embeddings live in, where position encodes relationships and distance approximates similarity.
Cosine similarity: A similarity measure based on the angle between two vectors; commonly used to compare embeddings, especially for text.
Nearest neighbor: The stored vectors closest to a query vector by a distance metric; the basis of vector search (often approximate / ANN at scale).
Centroid: The average point of a group of vectors — e.g. a seed-audience summary.
Vector database: A store optimized for indexing and similarity-searching embeddings at scale.
Semantic search: Retrieval by meaning — matching a query embedding to similar content embeddings rather than exact keywords.
Embedding portability: Whether embeddings from one model or system can be meaningfully compared or exchanged with another; usually limited without shared models or translation layers.
Translation layer: A mapping that aligns one embedding space to another so vectors can be compared across systems.
Projection layer: A learned transformation that maps vectors into a shared or lower-dimensional space for comparison.
Re-identification risk: The risk that data (including a vector) can be linked back to an individual or household using means reasonably likely to be used; central to whether embeddings are personal data.

Standards + agentic

AdCP: Ad Context Protocol. Open standard for agentic advertising — how buyer agents discover, query, and activate signals from sellers.
Advertising Context Protocol: The full name of AdCP: an open standard defining agentic advertising tasks and schemas over MCP and A2A transports.
AAMP: Agentic Advertising Management Protocols — IAB Tech Lab's umbrella initiative for agentic advertising standards across foundations, protocols, and trust.
ARTF: Agentic Real Time Framework — IAB Tech Lab's containerized real-time execution framework for agent services inside host platforms (v1.0 public comment).
Agentic Audiences: IAB Tech Lab's embedding-based signal-exchange standard (formerly UCP), donated by LiveRamp; identity, contextual, and reinforcement signals as 256–1024-dim vectors.
Agent Registry: IAB Tech Lab's trust-and-transparency registry for agent identity, verification, and disclosure (live March 2026).
A2A: Agent-to-Agent protocol — one of AdCP's two transports alongside MCP.
adagents.json: Publisher-side file declaring authorized AdCP agents and inventory; used for discovery and authorization.
brand.json: Buy-side machine-readable brand identity file in AdCP's Brand Protocol.
Agentic Direct: MCP-capable agentic version of OpenDirect.
Agentic Deals: MCP-capable agentic version of the Deals API.
Agentic Ad Objects: Agentic ad object model derived from AdCOM.
Agent evaluation: The methodology and tooling for testing whether a model or agent produces correct, useful, and safe outputs across approved tasks. Required before exposing agents to production data.
Agent-ready data: Data that carries governance, semantic context, evaluation, and feedback loops — not just access. Agentic workflows fail when data is available but semantically and governance-thin. Used in: Semantic Infrastructure Semantic & Agent-Ready
Buyer agent: An AI agent operating on the buyer side of an AdCP transaction. Issues spec-conformant requests to seller signal providers.
Contextual signal prototype: A semantic representation of an audience or signal concept, used to compare real-time bidstream context against the intended audience logic — the basis for identity-free, contextual activation.
Dual-gate activation: An activation pattern that makes an impression eligible only when both gates pass: audience membership AND contextual alignment. Reduces waste versus matching on either alone.
EARO: Exposure → Attention → Relevance → Outcome. Measurement framework connecting media exposure to business outcomes through attention + relevance. Used in: Multi-cloud data orchestration
Feedback loop: Captured user signals (corrections, ratings, escalations) that flow back into metric definitions, semantic mappings, and agent evaluation. Improves trust over time.
MCP: Model Context Protocol. An open standard (introduced by Anthropic, now governed under the Linux Foundation) for connecting AI assistants to external data and tools. It moves context between agents and systems; a semantic layer is still needed to supply shared meaning.
Observability: The instrumentation that lets a team see what data, queries, agents, and outputs are doing in production — traces, logs, metrics, alerts. Required for governed agentic workflows.
Portable intent: Buyer intent packaged so it can travel across platforms without being manually reinterpreted at each step. Also called a signal container or executable signal.
Signal container: A governed advertising signal object that carries the meaning, source, policy, activation path, and evaluation logic needed to move from buyer intent to platform execution. Also: executable signal, portable intent.
Signal containerization: Packaging buyer intent, semantic meaning, provenance, privacy policy, activation rules, and measurement logic into a portable advertising signal object that agents and platforms can discover, execute, and audit.
Tool catalog: The approved, allow-listed set of tools, APIs, and queries that an agent is permitted to call. Defines the operational boundary of the agent.
Tracing: Step-by-step record of an agent invocation — which tools it called, with what inputs, returning what outputs. The audit primitive for agentic systems.
UCP: Unified Context Protocol. IAB Tech Lab spec for agentic audience interoperability.

Core AdTech rails

OpenRTB: The IAB Tech Lab API specification for putting individual ad impressions up for bid in real time — the transaction rail beneath programmatic. The 2.x line is actively maintained (OpenRTB 2.6, with roughly monthly non-breaking updates); OpenRTB 3.0 exists as a separate spec line. Used in: Core AdTech Standards IAB Agentic Standards
AdCOM: Advertising Common Object Model. IAB Tech Lab’s shared object model for ads, placements, and context, used by OpenMedia specs including OpenRTB 3.0 and OpenDirect — so different protocols describe the same ad objects the same way.
OpenDirect: The IAB Tech Lab API spec for automating direct-sold / Automated Guaranteed inventory transactions — RFP, negotiation, and order management. Current version 2.1 (July 2024); uses AdCOM to describe media.
Deals API: IAB Tech Lab standard for one-way pushing of static deal terms from origin systems (SSPs, curation platforms) to receiving systems (DSPs), reducing manual deal entry. Version 1.0 was finalized February 6, 2026 — new enough that adoption should be validated, not assumed.
VAST: Video Ad Serving Template. The IAB Tech Lab XML schema for transferring video ad metadata from ad servers to video players. Current version is VAST 4.3 (December 2022), supplemented by the VAST CTV Addendum 2024. Used in: Video & Mobile Ad Delivery
VAST Wrapper: A VAST response that redirects the video player to another ad server for the actual creative, forming a chain of hops before anything renders. Each hop adds latency, error, and transparency risk — wrapper depth is a delivery variable worth auditing, especially in CTV.
Universal Ad-ID: The unique creative identifier carried in the UniversalAdId element of VAST 4.x so the same creative can be tracked consistently across platforms; Ad-ID is the registry for US markets. The VAST CTV Addendum 2024 retrofits UniversalAdId recognition into VAST 2.0–3.x via backward-compatible extensions.
VPAID: Video Player-Ad Interface Definition. Officially deprecated — IAB Tech Lab states it “has been deprecated and is being replaced by” OMID for measurement and SIMID for interactivity; the last published version is 2.0, so treat it as historical / legacy only.
SIMID: Secure Interactive Media Interface Definition. IAB Tech Lab’s standard for interactive video that separates the secure interactive layer from the media asset (both delivered via VAST 4.x), enabling SSAI and OTT use; current version 1.2. Together with OMID, the official replacement for VPAID. Used in: Core AdTech Standards
MRAID: Mobile Rich Media Ad Interface Definitions. The common API between rich media ads and the in-app environments they run in; version 3.0 (June 2018, backwards compatible) added viewability measurement support, audibility, a standardized close button, and error handling — it standardizes ad-to-app communication but does not solve measurement by itself.
SSAI (Server-Side Ad Insertion): Stitching ads into the video stream on the server before delivery — common in CTV and live streaming, and a source of verification and transparency considerations. The official IAB Tech Lab artifact is the SSAI VAST Macros Guidance v1.0 (2020) — guidance built on a curated VAST-macro subset, not a standalone spec — and SSAI should never be confused with SSAID, the Android device identifier. Used in: Core AdTech Standards
ads.txt: Authorized Digital Sellers. A public text file on a publisher’s domain declaring which companies are authorized to sell its inventory — a basic anti-spoofing rail. Version 1.1 (2022) added OWNERDOMAIN and MANAGERDOMAIN.
app-ads.txt: The extension of ads.txt to apps distributed through mobile and CTV app stores, so buyers can verify authorized sellers of app inventory. Version 1.0, final (2019).
sellers.json: A public JSON file ad systems publish so buyers can discover and verify the entities that are direct sellers of — or intermediaries in — the sale of digital advertising. Final specification dated July 31, 2019; works in tandem with the SupplyChain Object.
SupplyChain Object: An OpenRTB object (version 1.0) listing every party that sold or resold a given bid request as a chain of nodes. Usable as an extension with OpenRTB 2.5; part of the source object in OpenRTB 2.6 and 3.0.
Content Taxonomy: IAB Tech Lab’s common language for describing content, used for contextual targeting and brand-safety classification. Current version 3.1; the 3.x line expands to 1,500+ categories and 2.x is deprecated.
Audience Taxonomy: IAB Tech Lab’s standardized naming for audience segments across demographic, interest, and purchase-intent tiers, so different vendors describe the same segment the same way. Current version 1.1 (October 2020).
Data Transparency Standard: IAB Tech Lab’s “data label” — a disclosure schema for audience segments covering provenance, collection method, and recency. Public documentation lists version 1.1 (2021); validate current status before relying on it.

Privacy, consent + media quality

GPP: Global Privacy Platform. IAB Tech Lab’s transport protocol for moving privacy, consent, and consumer-choice signals from sites and apps to ad tech providers, with jurisdiction-specific sections (TCF EU, TCF Canada, the MSPA US National string, and individual US state strings). Used in: Privacy, Consent & Platform APIs Core AdTech Standards
TCF: Transparency & Consent Framework. IAB Europe’s GDPR consent framework, with technical specifications maintained by IAB Tech Lab. The operating version is v2.2; a v2.3 draft went through public comment (ended May 2025) but is not the operating version — validate current status.
Privacy Taxonomy: IAB Tech Lab’s universal classification language for data elements, data uses, and data subjects, giving businesses an interoperable way to label data for privacy work, data-subject rights, and partner exchange. No formal version number is published — validate current status.
Data Deletion Request Framework: IAB Tech Lab’s standardized mechanism for transmitting data deletion (“Right to Delete”) request signals through the ad supply chain — request packets, propagation, signatures, and response codes. Finalized in 2024; carries no version number.
Accountability Platform: An IAB Tech Lab specification (v1.0, finalized November 5, 2024) for open, auditable data structures that detect miscommunication of privacy preference signals such as GPP and TC strings across the supply chain. A specification, not a live operated service.
Privacy Sandbox: Google’s program of browser and Android advertising APIs designed as alternatives to cross-site tracking. Validation-sensitive: Google reversed third-party cookie deprecation (April 2025) and announced retirement of most Sandbox ad APIs (October 2025) with no published removal dates — validate current status before planning around it.
Topics API: A Privacy Sandbox API designed for interest-based advertising without sharing the specific sites a user visited. Google announced its retirement in October 2025 — treat as a winding-down browser API and validate current status.
Protected Audience API: A Privacy Sandbox API (formerly FLEDGE) designed to run on-device ad auctions in the browser for remarketing without cross-site tracking. Google announced its retirement in October 2025 — validate current status.
Attribution Reporting API: A Privacy Sandbox API designed to measure ad-interaction-to-conversion journeys via browser-generated reports instead of cross-site tracking. Google announced its retirement (Chrome and Android) in October 2025 — validate current status.
SKAdNetwork: Apple’s privacy-preserving install-attribution framework for iOS ad campaigns; version 4 supports up to three conversion windows (iOS 16.1+) and delayed, aggregated postbacks. Not deprecated as a framework — it coexists and interoperates with the newer AdAttributionKit. Validation-sensitive.
AdAttributionKit (AAK): Apple’s newer attribution framework (iOS 17.4+) for install and re-engagement attribution in the App Store and alternative app marketplaces, with cryptographically signed postbacks and no ATT authorization required. Interoperable with SKAdNetwork; validate current Apple documentation.
App Tracking Transparency (ATT): Apple’s consent framework: apps must request user authorization before accessing app-related data for tracking across other companies’ apps and websites. A hard platform constraint on mobile measurement; official documentation shows it remains in force.
IDFA: Identifier for Advertisers — Apple’s iOS advertising identifier, gated behind App Tracking Transparency authorization. Never assume it is available: without ATT permission an app cannot access it, and it is neither stable nor universal.
AAID (Android Advertising ID): Android’s advertising identifier (also called GAID). It is user-resettable, and since a late-2021 Google Play services update it is zeroed out when a user opts out of ads personalization; Google Play policy forbids connecting it to personally identifiable information or to persistent device identifiers (including SSAID) without consent.
App Set ID: A Google Play services identifier scoped to a developer’s own set of apps, which official Android documentation assigns to same-developer analytics and fraud-prevention use cases — not advertising. It resets under several conditions (for example after 13 months without access), so it should not be treated as persistent.
SSAID: SSAID, also known as Android ID, is a device identifier concept. It should not be confused with SSAI, server-side ad insertion, and should not be treated as an advertising delivery standard. On Android 8.0+, official documentation scopes it to each combination of app-signing key, user, and device.
MRC: Media Rating Council. The US industry self-regulatory body (established 1963 at the request of Congress) that sets minimum measurement standards and accredits measurement products through annual independent CPA audits. The trust anchor beneath viewability, IVT, and audience metrics. Used in: Measurement & Media Quality Research & Measurement Science
TAG: Trustworthy Accountability Group. A global industry initiative against fraud and criminal activity in digital advertising, operating four programs per its public documentation: Certified Against Fraud, Certified Against Malvertising, Brand Safety Certified, and Certified for Transparency.
Viewability: Whether an ad had the opportunity to be seen. Per MRC guidelines (August 2015, still the listed standard): a display impression is viewable at ≥50% of pixels in view for ≥1 continuous second; video at 50% of pixels for 2 continuous seconds.
Invalid Traffic (GIVT): General Invalid Traffic. Per MRC definitions, invalid traffic identifiable through routine, list-based filtration or standardized parameter checks — data-center traffic, known bots, spiders, crawlers, and non-browser user agents.
Sophisticated Invalid Traffic (SIVT): Per MRC definitions, invalid traffic that is harder to detect and requires advanced analytics, multi-point corroboration, and significant human intervention — for example app ID spoofing and domain laundering.
OMID: Open Measurement Interface Definition — the API at the heart of IAB Tech Lab’s Open Measurement program, defining how verification scripts collect measurement signals from ads in apps, web video, and (where supported) CTV; the current API version is 1.6, paired with the OM SDK 1.5 release line. Together with SIMID, the official replacement for VPAID. Used in: Measurement & Media Quality
Device Attestation: A capability IAB Tech Lab launched in the OM SDK on October 23, 2025 to combat device spoofing, with published implementation guidance and a public GitHub repo. Platform support varies by source — the launch announcement lists Apple devices and FireTV, while a later Tech Lab blog lists Samsung, LG, AndroidTV, and tvOS — and it is not a complete fraud solution; validation-sensitive.
CTV Verification: Independent confirmation that a CTV ad delivered as bought — harder than web verification because CTV inventory often runs through SSAI and closed app environments. OM SDK CTV support exists where platforms allow it (official pages list tvOS, Android TV, Samsung, and LG); validation-sensitive — validate current platform coverage before relying on it.
Brand Safety: The controls that keep ads away from broadly harmful or illegal content categories. The floor — distinct from brand suitability, which is brand-specific judgment above that floor.
Brand Suitability: Brand-specific judgment about which content is appropriate for a particular advertiser, layered on top of the brand-safety floor. The same content can be safe in general and unsuitable for one brand.
Verification: Independent, third-party confirmation that an ad delivered as bought — measured against viewability, invalid traffic, brand safety, and geography — rather than relying on a platform’s own reporting alone.
Media Quality: The umbrella for whether bought media was real, viewable, safe, and suitable — the layer where viewability thresholds, IVT filtration, verification, and accreditation programs together decide whether delivery can be trusted.

DOOH + place-based media

DOOH: Digital Out-of-Home. Digital ad screens in public and commercial spaces — billboards, transit, retail, venue networks — where one play typically reaches many viewers and the audience count is a modeled estimate, not a per-person log.
Digital Out-of-Home: See DOOH. The IAB definition (December 2024) frames it as dynamic digital screens combined with data-driven marketing, engaging one-to-one or one-to-many audiences in public and commercial spaces.
OOH: Out-of-home. The umbrella for advertising that reaches people outside their homes — billboards, street furniture, transit, place-based screens. DOOH is its digital, dynamically served subset; OAAA (founded 1891) is the US trade association for the category.
Place-based media: Screens and media tied to a specific venue — gyms, clinics, offices, retail, transit hubs — where the venue context is part of what is being bought. In programmatic DOOH the venue is described through a venue taxonomy rather than a URL.
Programmatic DOOH: Transacting DOOH through real-time bidding. There is no standalone “OpenRTB DOOH” spec — DOOH lives inside OpenRTB 2.6 (April 2022) via the dooh object (mutually exclusive with site and app) and the qty impression-multiplier object, plus an estimated-fulfillment timestamp (imp.dt). Used in: Core AdTech Standards
Proof-of-play: Evidence that a DOOH ad actually played on a screen. An operational concept, not a single cross-industry technical standard — reporting practice varies and self-verification remains common; the closest spec-level analogue is OpenRTB’s burl billing-event guidance (fired when the creative has rendered and impressions are billable).
Screen multiplier / Qty object: OpenRTB 2.6’s qty object passes a multiplier representing the total impressions for an ad that displays to more than one person. Values are statistically modeled and can be fractional (e.g. 0.32); cost resolves as (AUCTION_PRICE / 1000) × AUCTION_MULTIPLIER.
Venue taxonomy / OpenOOH: A standardized hierarchy of DOOH venue types for programmatic targeting. The OpenOOH Venue Taxonomy (spec 1.2.1, finalized February 2026) defines three tiers under 11 parent categories; documented production adoption sits on versions 1.0–1.1, and AdCOM recognizes multiple venue taxonomies (OpenOOH, DPAA, DMI).
Audience estimation: How DOOH audiences are counted: modeled estimates of how many people had the opportunity to see a screen, built from location and movement data. In the US, Geopath (founded 1933) is the audience measurement organization — treat counts as estimated, not observed.
Footfall attribution: Tying store visits to DOOH or other media exposure via location data. Correlated and modeled by default — a visit overlap is not a causal claim; causality needs experimental design such as control groups or holdouts.
Location-based attribution: The broader family of methods that use device location to link physical-world outcomes (visits, dwell) to ad exposure. Implementation-sensitive and consent-dependent; treat outputs as modeled estimates and validate the methodology before relying on it.

Audio + podcast advertising

Digital audio: Streaming music, internet radio, and podcasts as an advertising channel. Ad serving runs over VAST 4.x — audio support was integrated in VAST 4.1 (November 2018), which replaced the separate DAAST spec.
Podcast advertising: Ads in podcast content — integrated (“baked-in”), dynamically inserted, or sponsorship reads. Measured server-side under the IAB Podcast Measurement Technical Guidelines, where the measured unit is delivery (downloads), not confirmed listening.
VAST audio: Audio ad serving inside VAST: version 4.1 added an adType attribute on the Ad element (video / audio / hybrid) and annotated the spec for audio applicability. There is no separate audio ad-serving template anymore. Used in: Video & Mobile Ad Delivery
DAAST: Digital Audio Ad Serving Template. Legacy: IAB Tech Lab states DAAST “has been deprecated, and is being replaced by VAST” with the instruction to use VAST 4.1 or above. Treat any DAAST dependency as historical.
Dynamic Ad Insertion / DAI: Ads selected and inserted at the time a podcast file is requested, rather than recorded into the audio. There is no standalone IAB DAI spec — the official treatment is Section 4.3 of the Podcast Measurement Technical Guidelines, alongside integrated and sponsorship ads.
Podcast download: The measured unit of podcast delivery — and a qualified one: under the IAB guidelines a download counts only after filtering (at least one minute of content delivered, a 24-hour de-duplication window, IP plus user-agent listener identity, invalid-traffic exclusions). Delivery, not listening.
Podcast Measurement Guidelines: IAB Tech Lab’s Podcast Measurement Technical Guidelines — server-log-based rules for counting downloads, listeners, and ad delivery. Current version 2.2 (May 2024); a voluntary IAB Tech Lab compliance program audits implementations annually. Validate current status. Used in: Measurement & Media Quality
Host-read ad: An ad read by the show’s host, often baked into the episode audio. Valued for trust and audience relationship; harder to rotate, frequency-cap, and attribute than dynamically inserted spots.
Streaming audio: Session-based, server-streamed music and radio. Ad delivery behaves more like web media — real-time ad calls over VAST audio — than like podcasts, where the file is downloaded and listening happens off-server.
Audio attribution: Linking audio ad exposure to outcomes. Largely modeled — household-level IP matching, promo codes, vanity URLs, surveys — because podcast delivery is download-based and listening happens off-server. Delivery is not listening, and listening is not outcome.
Brand suitability in audio: Applying brand-suitability judgment to audio content — episode topics, host commentary, listener context — where transcripts and classification signals are less standardized than for web text. Implementation-sensitive; tooling coverage varies by publisher and hosting platform.

Prebid + header bidding

Prebid: Not a protocol standard — the open-source publisher execution layer where many standards become operational. Per its own docs, “a product suite, a community, and an independent organization”: free and fully open source (Apache license), overseen and funded by member ad-tech companies, with the technology available to non-members. Used in: Core AdTech Standards Privacy, Consent & Platform APIs
Prebid.js: The core product of the Prebid suite: a JavaScript library running in the browser (launched 2015) that calls selected bidders concurrently within a set timeout. Built as a custom bundle of modules — bid adapters, analytics adapters, user ID modules. Current line is 11.x with roughly weekly releases; validate current status.
Prebid Server: Prebid’s open-source server-to-server header bidding solution, with Go and Java implementations and 230+ server-side adapters. Official use cases: mobile app, AMP, server-side web with Prebid.js, and server-side ad inclusion scenarios such as CTV, digital out of home, and audio.
Prebid Mobile: Prebid’s open-source SDK for iOS and Android in-app header bidding (current line 3.0). It requires a Prebid Server — the auction runs server-side, not on-device — and supports four integration methods, including Google Ad Manager and mediation via AdMob and AppLovin MAX.
SharedID: Per current Prebid docs, “a convenient Prebid-owned first party identifier” within the User ID module framework — a random UUID stored in a first-party cookie on the publisher’s domain. Since Prebid.js 5.0 it makes no external sync calls, and reads and writes are suppressed without TCF Purpose 1 consent.
User ID module: Prebid.js’s framework for pseudonymous identity submodules (60+ listed, including SharedID, UID2, and ID5). Publishers configure storage per submodule and can restrict which bidders receive each ID; IDs surface to adapters in ORTB EID format, gated by consent modules.
Bidder adapter: The Prebid module that translates an auction into a specific demand partner’s bid request and response. Prebid’s docs cite more than 300 demand partners client-side; each publisher compiles only its selected adapters into its Prebid.js bundle.
Analytics adapter: A vendor-specific Prebid.js module, compiled into the bundle at build time and enabled at runtime via enableAnalytics(), that hooks into the Prebid event system to capture auction and bid performance.
Price floors: The lowest CPM a bid must meet in a Prebid auction, set through the Price Floors module — at ad-unit level, via setConfig, or fetched dynamically from a floor provider. Prebid’s own docs discourage static floors as blunt and quickly stale.
Dynamic floors: Floors fetched at auction time from a floor-provider endpoint instead of hard-coded values. Floors schema 2 lets providers A/B-test floor groups through weighted model sampling — the approach Prebid’s documentation favors over static configuration.
ORTB2: Prebid’s OpenRTB-shaped convention for first-party data: setConfig({ortb2}) carries site, user, and content objects (with segtax-labeled taxonomy segments) to all bidders; setBidderConfig() scopes data to named bidders; ortb2Imp carries ad-unit-level data.
Client-side header bidding: Auctions run in the user’s browser via Prebid.js: bidders are called directly from the page within the timeout. Direct access to browser context where permitted, at the cost of page weight and latency as bidders are added.
Server-side header bidding: Moving auction calls off the page: one request goes to Prebid Server, which fans out to bidders server-side. Lighter pages and room for more bidders, with implementation-sensitive trade-offs in cookie and ID availability.
Auction timeout: The window Prebid gives bidders to respond before the auction closes and the ad server is called. The central latency-versus-revenue tuning knob: too short drops bids; too long delays the page.
Bid density: How many bidders return valid bids per auction — an operational health metric for a header-bidding setup, not a standard. Low density points to misconfigured adapters, over-tight timeouts, or floors set above what demand will pay.
Bidder permissions: Prebid’s controls over which bidders receive which data: setBidderConfig() scopes first-party data to named bidders, the User ID module’s bidders array restricts ID distribution, and Activity Controls (allowActivities) gate activities such as transmitting user FPD or EIDs. Mechanics inside Prebid.js — not a privacy-compliance guarantee.
Publisher control layer: The operating frame for Prebid’s role: the layer where the publisher decides which bidders compete, what data and IDs they receive, what floors apply, and how consent signals gate activity. Execution control — not a standards body, not an SSP or DSP, and not a replacement for the ad server.

Retail + commerce media measurement

In-store media: Advertising delivered inside the physical store — digital screens, audio, and connected shopping formats. The IAB / IAB Europe In-Store Retail Media: Definitions and Measurement Standards (final, December 2024) define five store zones — Exterior; Entrance & Out of Category; Check out; In Aisles; Other & Connected Store — plus the official media, traffic, and impression terms. Other formats are deferred to future iterations.
Closed-loop attribution: Assigning credit for sales to specific ad exposures using the retailer's own transaction data. The IAB/MRC Retail Media Measurement Guidelines (January 2024) define attribution as assigning credit for consumer actions to specific marketing efforts — and note that MRC requires viewable impressions for attribution of outcomes, something the guidelines acknowledge most retail media organizations do not yet do.
SKU-level measurement: Reporting media outcomes at the individual item (stock-keeping-unit) level — which products sold, not just whether a campaign converted. A core retail media promise; definitions and match logic vary by network, so cross-network comparability has to be verified rather than assumed.
Basket analysis: Analyzing what else was purchased alongside an advertised product — basket size, attach, and category effects. Useful for halo and category questions; an analysis practice defined per network rather than a standardized metric.
New-to-Brand: Buyers who had not purchased the brand within a lookback window. One of the optional metric definitions in the IAB/MRC Retail Media Measurement Guidelines (January 2024) — explicitly not mandatory for reporting — and, with New to Category, defined in the December 2024 in-store standards with purchase-cycle timeframes. Lookback windows vary by network.
Sales lift: The change in sales associated with a campaign versus a baseline or control. The December 2024 in-store standards define Sales Lift alongside Brand Sales Lift (Halo — same brand, same category) and Incremental Sales measured test-versus-control. A lift number is only causal when the control is credible.
Ad Play: Official in-store impression term (IAB / IAB Europe, December 2024): the number of times an ad is displayed or rendered on a particular format. The base delivery count — it says nothing about whether anyone was present to see it.
Gross Impression: Official in-store impression term (December 2024): the number of individuals present in the defined Display Exposure Zone while the display is functional — ad impression = audience × ad play. Presence, not confirmed viewing.
Opportunity To See (OTS): Official in-store impression term (December 2024): a single opportunity to view an advert — the in-store analogue of a viewable impression. The standards position OTS as the best available in-store proxy for a viewable ad impression for digital activations.
LTS Impressions: Likelihood-To-See — official in-store impression term (December 2024): a refinement of viewable impressions that adjusts for the likelihood individuals actually noticed the ad, typically determined using sensor or analytic technology. The most modeled tier of the in-store impression ladder.
Verified impression: Not an official term: it does not appear in the IAB/MRC Retail Media Measurement Guidelines (January 2024) or the IAB / IAB Europe in-store standards (December 2024), whose impression terms are Ad Play, Gross Impression, Opportunity To See, and LTS Impressions. Where vendor or trade material uses "verified impressions," ask which official metric it maps to.

Clean rooms, PETs + interoperability

OPJA: Open Private Join and Activation. IAB Tech Lab's first clean-room interoperability standard for privately activating audiences — now deprecated and no longer supported, superseded by PAIR after Google donated that protocol to Tech Lab. Reference historically only.
Secure matching: Joining two parties' records on encrypted or hashed keys so neither side sees the other's raw identifiers — for example PAIR's commutative encryption (keys are multiply-encrypted and matched without decryption) or private-set intersection. It reduces exposure during the match; it does not by itself govern what outputs may leave.
Clean room activation: Turning a clean-room match into live targeting or suppression. PAIR's pattern: offline matching inside a data clean room, then programmatic activation with encrypted publisher IDs carried in OpenRTB's eids object. Activation rights are scoped separately from the right to analyze.
Clean room measurement: Attribution and measurement run across parties' data inside a governed environment. IAB Tech Lab's ADMaP v1.0 — the Attribution Data Matching Protocol, finalized February 2025 — is the clean-room interoperability standard for attribution measurement, using private-set intersection and trusted execution environments.
Audience overlap: The count or share of individuals present in both parties' datasets — typically the first governed query a clean-room collaboration runs. Outputs are gated by aggregation thresholds so small overlaps cannot single anyone out.

Event + conversion APIs

ECAPI: IAB Tech Lab's Event & Conversion API. Version 1.0 was finalized in 2026 (public comment January–February; declared available for implementation in April) and standardizes server-to-server transmission of marketing events from advertiser systems to platforms and partners — a named event taxonomy, user identifiers, value and currency, and GPP-based consent metadata. A shared layer over platform CAPIs, not a replacement; new enough that platform support must be validated.
Conversion API (CAPI): A platform's server-to-server endpoint for receiving conversion and marketing events directly from an advertiser's systems rather than from browser tags. Each platform operates its own, with differing structures and requirements — the per-platform integration burden ECAPI was created to standardize.
Server-to-server events: Marketing events sent from an advertiser's server to a partner's API endpoint instead of fired client-side. More resilient to browser and app signal loss; consent obligations and deduplication against client-side signals still apply.
Outcome signal: A conversion or business event — purchase, lead, subscription — fed back to a buying or optimization system as the thing to optimize toward. Quality, deduplication, and consent metadata determine whether the signal is safe to optimize against: agents should not optimize against outcomes they cannot trust, interpret, or audit.
Event taxonomy: A named, standardized set of event types so every party means the same thing by an event. ECAPI 1.0 defines standard events spanning the funnel — purchase, page_view, ad_impression, add_to_cart, begin_checkout, generate_lead, sign_up, install, subscribe, and more, plus a custom type — rather than each platform inventing its own names.
Deduplication key: The identifier used to count an event exactly once when it arrives by more than one path. ECAPI deduplicates on the unique combination of data_set_id and event id; Meta deduplicates Pixel and Conversions API events on a shared event_id. Without a dedup key, server-side event adoption inflates counts.
Offline conversion: A conversion completed outside the measured digital surface — an in-store purchase, a phone-closed deal, a CRM stage change — uploaded to platforms server-side so optimization and measurement see outcomes the pixel never could. Match rates and consent determine how much of it lands.
Consent metadata: Machine-readable consent context carried with an event. ECAPI 1.0 carries GPP fields — gpp_string and gpp_sid — plus an mmt_only flag marking events for measurement only, excluded from ads-optimization use. Carrying the metadata does not settle legal questions; the spec leaves that responsibility with implementers.
Full-funnel events: Sending upper-, mid-, and lower-funnel events — views, carts, checkouts, purchases, leads — rather than purchases alone, so optimization and measurement see the path and not just the endpoint. ECAPI's taxonomy is explicitly built for events across the funnel, not just conversions.

CTV, streaming + live events

CTV Ad Portfolio: IAB Tech Lab's portfolio of six core CTV ad formats — Pause Ad, Menu Ad, Screensaver Ad, In Scene Insertion Ad, Overlay Ads, and Squeeze Back Ads — distilled from 100+ industry submissions. Announced December 2025; public comment closed January 31, 2026, with companion signaling guidance still in comment in mid-2026 — validate current status before treating any of it as final.
Pause Ad: An ad shown when the viewer pauses content — one of the six CTV Ad Portfolio formats, and one of the two (with Menu Ads) prioritized for programmatic transaction support in the updated Guide to Programmatic CTV (December 2025). Portfolio status: public comment closed, finalization unverified — validate before relying on it.
Live event ad serving: Ad delivery for live streams, where millions of viewers can hit the same ad break at once. IAB Tech Lab's Live Event Ad Playbook (LEAP, August 2025) collects technical recommendations and protocol enhancements for low-latency, reliable live ad delivery, developed with collaborators including Amazon, NBCUniversal, FreeWheel, and Index Exchange — collaboration on the standard, not verified production deployment.
Concurrent Streams API: IAB Tech Lab specification (v1.0, final — public comment closed July 2025) standardizing how subscribers query the number of viewers on a live stream in real time, so ad decisioning can scale ahead of concurrency spikes instead of dropping ad breaks. The shipped LEAP component; others (Forecasting API, Buyer Instructions API) were still in comment or expected as of mid-2026.

Sustainable media + carbon measurement

Sustainable media: Planning and buying media with greenhouse-gas emissions treated as a managed variable alongside price and performance — using voluntary frameworks such as the GMSF and IAB Tech Lab's Sustainability Playbook. The frameworks produce estimates and practices, not certifications; no buy is green by default.
Media carbon measurement: Estimating the greenhouse-gas emissions attributable to media activity. Outputs are estimates built on emissions factors and assumptions, not metered readings — the GMSF's own data-gap mechanisms explicitly reduce accuracy when inputs are missing. Treat numbers as comparable decision inputs, not audited facts.
GMSF: Global Media Sustainability Framework. Voluntary industry standards for consistent, comparable greenhouse-gas estimates across digital, TV, print, audio, out-of-home, and cinema. Launched June 2024 by WFA's GARM with the Ad Net Zero community; published and stewarded by Ad Net Zero since GARM was discontinued (August 2024). Current version v1.2 (June 2025), aligned with the GHG Protocol and ISO 14067 — validation-sensitive, with a v1.3 announced but not verified as shipped.
Supply-path emissions: Emissions attributable to how programmatic moves data — duplicated bid requests, multi-hop resale, cookie syncs. IAB Tech Lab's Sustainability Playbook (June 2023) frames the lever directly: reducing carbon in programmatic means reducing the requests and data the ecosystem processes. Best-practices guidance, not a measurement methodology.
Programmatic sustainability: Cutting waste in programmatic delivery — fewer reseller hops, deduplicated supply paths, pod bidding, Global Placement ID adoption — so the same outcome ships with fewer requests processed. The efficiency case and the emissions case largely point the same way; specific savings claims still need validation.
Greenwashing risk: The risk of presenting voluntary, estimate-based sustainability outputs as verified environmental claims. GMSF outputs are voluntary estimates; even advertising's share of global emissions is framed by Ad Net Zero as a rough 2–4% that cannot be pinpointed. Treat carbon-neutral media claims as marketing until the methodology is shown.

Research + measurement science

ARF: Advertising Research Foundation. An industry research body advancing the scientific practice of advertising, marketing, and media research — councils, effectiveness studies, original research, and best practices. A research / evidence layer, not a protocol.
Advertising Research Foundation: See ARF. The industry research body behind the Journal of Advertising Research; it integrated MSI (2021) and acquired CIMM (2018).
MSI: Marketing Science Institute. An academic–practitioner bridge for marketing science and periodic research priorities; a division of the ARF since 2021.
Marketing Science Institute: See MSI. Connects academic marketing science to business practice and sets periodic research priorities (analytics, AI, measurement, customer journeys, innovation).
CIMM: Coalition for Innovative Media Measurement. Industry-backed media-measurement innovation across cross-platform and TV / CTV measurement, currency, and identity; acquired by the ARF in 2018.
Coalition for Innovative Media Measurement: See CIMM. Especially relevant to converged TV, streaming, CTV, identity, and currency questions.
JAR: Journal of Advertising Research. The ARF’s peer-reviewed, practitioner-facing journal (published since 1960) — the deeper evidence layer behind advertising effectiveness and methodology.
Journal of Advertising Research: See JAR. Peer-reviewed advertising research published by Taylor & Francis on behalf of the ARF; cite public metadata and summaries, not member-only content.
Measurement science: The discipline of establishing which advertising evidence is valid, causal, and comparable — distinguishing exposure, attention, persuasion, behavior, incrementality, and business value rather than collapsing them.
Marketing effectiveness: Whether advertising produced trusted, incremental business outcomes — not just delivery or platform-reported activity.
Cross-platform measurement: Comparing reach, exposure, and outcomes across linear, streaming, CTV, digital, and retail media with consistent definitions. Inconsistent definitions create planning errors.
Currency (measurement): The agreed measurement standard used to plan and transact media. Comparability and currency are central CIMM questions for converged TV and video.
Attention measurement: Evidence that an exposure received human attention. Attention is not the same as persuasion or a sales effect; it needs validation against memory or outcome.
Causal inference: Methods — randomized experiments, holdouts, geo tests, marketing mix modeling — for establishing what an activity actually caused, versus correlation captured by platform attribution.
Research validity: Whether a measurement actually measures what it claims, free of biased samples, weak proxies, and poor practice — a core risk for AI-assisted and synthetic-data research.
Evidence hierarchy: The ordering of evidence strength — from exposure and attention up through incrementality and business value — that agentic systems must respect rather than collapse into one metric.

Platforms & fit

Data gravity: The tendency of workflows to move toward where the data already lives. The buyer's existing data estate often decides the platform before capability does. Used in: Multi-cloud orchestration Snowflake
Secure data sharing: In-place sharing of governed objects without copying data between accounts. Distinct from a clean room — sharing grants access; a clean room governs joint analysis.
Native app: Application logic that runs next to the consumer's data so the provider's IP and code stay protected. A Snowflake packaging option beyond a raw share.
Marketplace listing: Public discovery + self-serve distribution surface for a data product, app, or model. Reduces distribution friction; not a GTM strategy on its own.
Private listing: A named-buyer, custom-terms distribution of a governed data product. Higher margin, slower cycle than a public marketplace listing.
Unity Catalog: Databricks' unified governance + discovery layer for data and AI assets across formats and engines. Often framed as the governance and discovery spine of a lakehouse.
Delta Sharing: Open, cross-platform data sharing without copying or ETL between estates. Vendor-neutral protocol associated with the Databricks ecosystem.
AI/BI: Governed dashboards plus conversational analytics over a curated semantic layer. Only as trustworthy as the metadata and metric definitions behind it.
AWS Clean Rooms: AWS-native, privacy-enhancing collaboration on sensitive data without sharing raw rows. Governed by configured tables and analysis rules. Used in: AWS platform fit Amazon Marketing Cloud
Analysis rules: Configured-table rules in AWS Clean Rooms defining which queries, joins, and aggregations a collaboration permits. The output-control primitive.
BigQuery clean rooms: BigQuery data clean rooms (generally available since April 2024, built on BigQuery sharing — formerly Analytics Hub): two parties analyze combined data in place, without copying or moving it, under analysis rules such as aggregation thresholds, join restrictions, and differential privacy. Distinct from Ads Data Hub — different tool, different rules.
Multi-cloud orchestration: An operating model — not a product — that decides where data lives, where logic runs, what outputs are allowed, and which platform owns each decision when a brand uses several clouds, clean rooms, BI, and activation systems.
Collaboration control plane: The cross-platform operating layer: use-case registry, ownership, identity/consent/join rules, semantic + metric definitions, output policy, activation rights, audit/lineage, monitoring, and feedback. Without it, every clean room and dashboard becomes a one-off project.
Platform role map: Assigning each platform a role by decision type (distribution, lakehouse intelligence, infrastructure-native collaboration, media measurement) rather than picking one platform to "win."
Compute-to-data: Moving the query, model, or application to where sensitive data already lives, instead of moving raw data out — the default in governed multi-cloud collaboration.
Logic-to-data: The pattern of moving governed functions, clean room queries, native apps, or model containers to the data rather than exporting raw rows. Move logic and approved outputs before moving data.
Activation rights: The explicit, contracted permission to use a collaboration output to activate — suppress, target, optimize, exclude. Separately scoped from the right to analyze; an output can be measurable but not activatable.
InfoSum: A neutral data collaboration platform organized around non-movement of data: two parties match and analyze against each other's records without pooling or moving raw data. Best assessed against a specific two-party match-and-measure decision. Validate current capabilities against official documentation. Used in: LiveRamp Platform fit
Non-movement of data: A collaboration model where raw records never leave each party's environment; only governed match logic and approved aggregate outputs cross the boundary. A hard requirement in many neutral clean rooms.
Private-set intersection (PSI): A cryptographic technique that lets two parties compute the overlap between their datasets without revealing the non-matching records. A privacy-enhancing primitive behind several non-movement clean rooms. Used in: Data Clean Rooms, PETs & PAIR
Governed join and compute: Running a join and an analysis across two parties' data under explicit rules — which keys may match, which queries may run, and what may leave — without either side seeing the other's raw rows.
LiveRamp: An identity resolution and data collaboration layer used to match and connect records across partners, clouds, and clean rooms (the LiveRamp Clean Room, Habu-powered) via a persistent identifier (RampID). Treated here as connective infrastructure, not a decision in itself. Validate current capabilities against official documentation. Used in: InfoSum CDP / DMP / First-Party Data
RampID: A persistent, people-based identifier used to resolve and connect records across environments — the interoperability anchor in LiveRamp-based workflows. Confirm current naming and availability against official documentation.
Pseudonymized identifier: A stable token that stands in for a person or record so data can be matched across parties without exposing direct identifiers. The unit of interoperability in most identity-based collaboration.
Amazon Marketing Cloud (AMC): Amazon's walled-garden clean room for measurement and audience analysis across Amazon Ads and Amazon DSP signals, with first-party data matched for analysis inside the environment. Scoped to Amazon media questions. Validate current capabilities against official documentation. Used in: Retail Media Network AWS
Retail media clean room: A walled-garden environment operated by a retailer or marketplace where advertisers analyze and activate against that retailer's commerce and media signals without receiving raw data. Amazon Marketing Cloud is the most prominent example.
Walled-garden clean room: A clean room owned and bounded by a single media ecosystem. Powerful for that ecosystem's own questions; not a neutral, cross-platform measurement layer.
Audience activation: Turning a collaboration output into a usable audience — built, suppressed, or targeted — inside an activation environment. Scoped separately from the right to analyze; an output can be measurable but not activatable.
Aggregation threshold: The minimum number of records or people a clean room result must represent before it can be returned or exported — a core output control in walled-garden and neutral clean rooms alike. Confirm the exact threshold per platform.
k-anonymity (k-min): A privacy threshold that suppresses any result representing fewer than k people, so small groups cannot be singled out. The minimum-audience or minimum-result rule in clean room output policy — LiveRamp calls this "Crowd Size"; Amazon Marketing Cloud applies built-in aggregation thresholds. Confirm the exact threshold per platform.
Differential privacy: A privacy-enhancing method that adds calibrated noise to query outputs so individuals cannot be re-identified from results. Applied by some collaboration environments to released outputs; confirm the exact controls per platform. Used in: Data Clean Rooms, PETs & PAIR
Privacy-enhancing technology (PET): Umbrella term for techniques that allow analysis on sensitive data while limiting exposure — clean rooms, private-set intersection, differential privacy, k-anonymity, encryption-in-use, and non-movement architectures. Used in: Data Clean Rooms, PETs & PAIR
Advanced Marketing Analytics (AMA): A cloud marketing-analytics operating model: connect data collection, transformation, analysis, measurement, modeling, and activation. The durable idea; the current environment is GA4, BigQuery, Ads Data Hub, clean rooms, Meridian, and privacy-safe activation.
GA4 BigQuery export: Google Analytics 4's raw, event-level export to BigQuery (daily and/or streaming). Free standard properties are capped at 1M events/day on the daily export; Analytics 360 raises the limits. Universal Analytics is retired (stopped processing in 2023).
BigQuery ML: Train and run ML models directly in BigQuery with SQL — no data movement — across model types such as logistic regression, boosted trees (XGBoost), k-means, and ARIMA forecasting.
Meridian: Google's open-source, Bayesian causal marketing-mix model (MMM) for privacy-durable measurement — aggregate data, no cookies or user-level IDs. A self-hosted framework you run yourself (the supported successor to the discontinued LightweightMMM), not a managed Google product.
Privacy checks: Ads Data Hub's built-in restrictions that block user-level export and enforce aggregation before results leave Google's cloud project. The default mechanism is noise injection (~20 unique users per row); difference checks (~50) are a legacy alternative.
Filtered row summary: An Ads Data Hub output that recaptures the aggregate of rows dropped by privacy checks (within privacy limits), so suppressed rows do not silently distort a total.
Marketing analytics pilot: A narrow six-week pattern — define decision → select use case → prepare data → build analysis → activate → production path — that turns one business question into one governed output before scaling.
Customer journey analytics: Analysis of how customers move across touchpoints (site, app, media, CRM, offline) to explain what happened and where value comes from. The "understand" rung of the marketing-analytics ladder.
Predictive audience: An audience defined by a model score (propensity, value, churn risk) rather than a fixed rule — built in the analytics layer and activated only where eligible.
Purchase propensity: A model score estimating how likely a customer is to purchase — a "predict" use case feeding ranked audiences and planning inputs.
LTV prediction: Modeling a customer's expected lifetime value to prioritise acquisition, retention, and high-value-customer discovery — the forward, modeled estimate (distinct from the CLTV / LTV definition).
Meta Advanced Analytics: A Meta-specific, access-controlled, privacy-safe measurement and analytics capability for deeper questions than Ads Manager answers (pathing, overlap, lift, custom attribution). It is not a self-serve product and is commonly reached via approved partners (e.g. LiveRamp, dentsu) — distinct from the discontinued consumer "Facebook Analytics." Validate current access and support. Used in: Google / GMP / ADH BI / MMM / Decision Intelligence
Conversions API (CAPI): Meta's server-side event-sharing API that complements the browser Pixel, recovering signal lost to ATT and browser privacy. Events are deduplicated against the Pixel using a shared event_id. Used in: Event & Conversion APIs
Conversions API Gateway: A code-free, self-serve setup option for the Conversions API, run from Events Manager — a faster path to server-side events without a custom build.
Aggregated Event Measurement (AEM): Meta's privacy-aware measurement for iOS 14.5+ users, with an 8-event-slot limit per domain and configurable event priority. It constrains and models what web/app conversions can be reported.
Event Match Quality (EMQ): A current Events Manager metric (Poor / OK / Good / Great) showing how well your events match to Meta accounts. Weak EMQ degrades optimization and measurement.
event_id deduplication: Sending a shared event_id on both the Pixel and Conversions API so Meta counts a single conversion once. Without it, server and browser events cannot be deduplicated.
Advantage+: Meta's family of AI / automated delivery products — including Advantage+ sales campaigns (the successor naming for what was Advantage+ Shopping / ASC), plus Advantage+ audience, creative, app campaigns, and catalog ads.
Conversion Lift: Meta's incrementality test (within the Experiments umbrella, alongside A/B tests): a test group vs a holdout, where the performance difference is the incremental lift caused by advertising. Use for causal read-outs, not last-touch attribution.
Ads Insights API: The current programmatic interface (part of Meta's Marketing API) for retrieving ad reporting and statistics — reporting data, not deeper analytics.
Robyn: Meta Marketing Science's open-source, Bayesian marketing-mix model (MMM) for privacy-friendly cross-channel budget allocation, response and saturation curves, and scenario planning. Active (not deprecated) but slowing release cadence — verify the current version and support before production.
MMM calibration: Using lift / experiment results as ground truth to calibrate a marketing-mix model, so MMM and platform measurement agree rather than living in separate silos.
SKAdNetwork (SKAN): Apple's privacy-preserving attribution framework for iOS app campaigns. Meta supports it, with campaign-level aggregated, delayed reporting.
Privacy-safe attribution: Attribution built on aggregated, consented, privacy-constrained signals (server-side events, modeled conversions, aggregation thresholds) rather than unrestricted user-level tracking.
Partner-mediated clean room: A walled-garden analytics capability reached through an approved partner rather than a self-serve UI — e.g. Meta Advanced Analytics via LiveRamp Clean Room or dentsu. Access, terms, and data windows are partner- and permission-dependent.

Ecosystem surfaces

Ecosystem surface: An execution surface where a governed data output becomes activation, measurement, optimization, or an agentic decision — CDP, DSP, retail media, publisher / SSP, BI / MMM, or semantic infrastructure. The surface is not the strategy; the semantic and governance layer that runs across surfaces is.
CDP: Customer Data Platform. The control plane for first-party customer data — ingest, identity resolution, consent, segmentation, and activation. Most valuable when it closes a measurement loop, not just produces segments. Used in: DSP / Agentic Buying LiveRamp
DMP: Data Management Platform. The third-party-audience tool of the cookie era, largely displaced as third-party identifiers declined — though DMP-style taxonomy and audience management persist inside CDPs, DSPs, and clean rooms. Used in: CDP / DMP / First-Party Data
DSP: Demand-Side Platform. Where governed audiences, signals, budgets, and goals become executable media decisions across display, video, CTV, audio, and the open web — increasingly via automated, agentic optimization. Used in: Google / GMP / ADH Retail Media Network
Agentic buying: A buying model where the operator sets the outcome, budget, and guardrails and the system plans, bids, optimizes, and measures. The risk is autonomous optimization against the wrong metric.
CTV: Connected TV. Premium streaming inventory bought programmatically; needs household identity and frequency control rather than display tactics. Used in: CTV, Streaming & Live Events Video & Mobile Ad Delivery
Retail media network (RMN): A retailer or marketplace selling access to its first-party purchase data as media — on-site, off-site, and in-store. The draw is closed-loop measurement against real sales; the unsolved problem is comparability across networks. Used in: Retail / Commerce Media Measurement Amazon Marketing Cloud Performance & Native
Closed-loop measurement: Tying ad exposure to actual purchases inside the environment that owns the sales data. Powerful, but defined differently by each retail media network — verify it before trusting the ROAS. Used in: Retail / Commerce Media Measurement Retail Media Network
Commerce media: The extension of retail media beyond grocery and mass retail to marketplaces, delivery, travel, and financial services — any platform with logged-in, purchase-level first-party data to sell against. Used in: Retail / Commerce Media Measurement
SSP: Supply-Side Platform. The sell-side of programmatic — where publishers run auctions and, increasingly, package first-party data, identity, content, and quality into curated supply. Used in: DSP / Agentic Buying
Curation: Packaging audience, identity, content, and quality into addressable supply on the sell side. The new control point for what data travels with the impression — but it can also rebundle weak supply behind a premium label. Used in: DSP / Agentic Buying
Supply-path optimization (SPO): Buyers consolidating to fewer, cleaner, more transparent paths to inventory. Rewards transparency; hidden fees, duplicate paths, and reseller / made-for-advertising supply get cut.
Header bidding: A sell-side technique that runs parallel auctions for an impression before the ad-server call, so demand competes simultaneously rather than through the legacy sequential waterfall. Prebid is the open-source execution layer built around it. Used in: Publisher / SSP / Curation
Curated Audiences (SDA): Publisher-declared audience segments for cookieless addressability. The IAB Tech Lab standard was renamed from "Seller-Defined Audiences" to "Curated Audiences"; real-world adoption has been uneven.
Marketing mix modeling (MMM): Aggregate, causal modeling of how channels drive outcomes — privacy-durable budget allocation that survived the decline of user-level tracking. Only as good as the experiments it is calibrated against. Used in: Google / GMP / ADH Meta Advanced Analytics
Incrementality: The causal contribution of an activity — what would not have happened without it — measured with geo tests, holdouts, or lift studies. Distinct from attribution, which assigns credit to conversions that may have happened anyway. Used in: BI / MMM / Decision Intelligence Performance & Native
Multi-touch attribution (MTA): User-path-based credit allocation across touchpoints. Granular and familiar, but losing reliability after privacy changes and losing primacy to causal and geo-based methods.
Decision intelligence: The layer that turns BI outputs into budget and growth decisions — increasingly reframed as "agentic analytics." Copilots are here; fully autonomous analysts mostly are not.
Agentic analytics: AI assistants that query, monitor, and propose actions over governed data. Natural-language copilots are largely generally available and human-supervised; autonomous multi-step analysts remain mostly emerging.
Semantic infrastructure: The universal navigation layer — shared definitions, identity, metadata, ontology, and governance — that lets fragmented data ecosystems become usable by humans, platforms, and AI agents.
Universal navigation layer: The framing of semantic infrastructure as the membrane between fragmented data (each silo speaking its own language) and the humans, platforms, and agents that need one trustworthy answer.
Context layer: The 2026 reframing of the semantic layer for the AI era — meaning plus governance exposed to agents. Largely the same idea as the semantic layer, relabeled.
Metrics layer / headless BI: Pulling metric definitions out of any single BI tool so every tool — and every agent — shares one definition. The portable core of semantic infrastructure.
Knowledge graph: A model of entities and their relationships that makes meaning explicit — used to ground LLMs and reasoning (GraphRAG) rather than leaving relationships implied in tables.
Open Semantic Interchange (OSI): A vendor-neutral, open specification for portable metric and semantic definitions, so meaning is not locked to one platform. Part of the standardization of the semantic layer.

App growth + mobile measurement

Apps DSP: A reframe of the mobile DSP: an app-growth decision layer spanning install, engagement, retention, LTV, commerce, and CTV extension — not just a mobile bidder. Describes the buyer problem, not the channel.
SKAdNetwork (SKAN): Apple’s privacy-preserving install-attribution framework for iOS; current version is SKAN 4 (no "SKAN 5" exists). Postbacks, fine/coarse conversion values, and lock windows replace deterministic, user-level attribution.
AdAttributionKit: Apple’s newer attribution framework (introduced WWDC 2024), built on SKAN’s foundations and adding re-engagement attribution and support for alternative app marketplaces. As of mid-2026 it coexists with SKAN — Apple has announced no forced cutover. Validate current Apple documentation.
MMP (mobile measurement partner): A neutral attribution + analytics layer for apps — AppsFlyer, Adjust (AppLovin-owned), Singular, Kochava, Branch, Airbridge — covering install, event, cohort, ROAS, retention, and fraud controls.
Conversion value schema: The design of SKAN / AdAttributionKit conversion values (fine 0–63 or coarse low/med/high) that maps post-install events into the privacy-limited postback. The single most consequential measurement-design choice for an iOS app campaign.
Geo-lift: A geo-based incrementality test — holding out matched regions to measure the net-new effect of spend (vs. attribution, which allocates credit to conversions that may have happened anyway). "GeoLift" also refers to Meta’s open-source library.
MMM for apps: Marketing mix modeling applied to app growth — reconciling spend, channel mix, saturation, seasonality, creative, organic lift, and CTV exposure against app outcomes. Complementary to (not a replacement for) MMP attribution and incrementality.
CTV app-growth bridge: The measurement path connecting CTV exposure back to an app outcome: exposure → identity bridge (graph / IP / login) → app or web action (QR / deep link) → app event → MMP / MMM / BI → budget and frequency decision. CTV only counts for app growth when this bridge holds.
Axon (AppLovin): AppLovin’s AI/ML engine (AXON) and, since Oct 2025, the umbrella brand for its advertising platform plus the self-serve Axon Ads Manager; AppDiscovery is the UA product. Having divested its games business in 2025, AppLovin is now a pure ad platform. Cited as category gravity, not a ranking.
LTV optimization: Optimizing acquisition and engagement toward predicted lifetime value (cohort LTV, payer quality, subscription revenue) rather than install volume or last-click CPI. The decision the apps-DSP frame is built around.
ATT (App Tracking Transparency): Apple’s consent prompt governing cross-app/website tracking and access to the IDFA. In force globally as of mid-2026 (tightened in iOS 26); its EU future is contested under regulatory pressure — monitor.
Android Privacy Sandbox (status): In October 2025 Google retired the core Privacy Sandbox advertising APIs (Attribution Reporting, Topics, Protected Audience) on both Chrome and Android. GAID and the Play Install Referrer remain the working backbone of Android app attribution; treat any roadmap claim as live risk and validate current official Android documentation.
Server-side / S2S events: Conversion and event data sent server-to-server (CAPI-style) rather than only client-side — de-risking signal loss from ATT, SKAN, and Android changes, and improving match quality across iOS and Android. Used in: Event & Conversion APIs
App re-engagement: Reactivating existing or lapsed users (retargeting / re-attribution) rather than acquiring new installs — a long-standing SKAN gap that AdAttributionKit begins to address.
App install attribution: Connecting an app install to the ad exposure that drove it — increasingly modeled and privacy-limited (SKAN / AdAttributionKit on iOS; GAID / Play Install Referrer on Android) rather than deterministic.
Cohort ROAS: Return on ad spend measured by install cohort over time (e.g. D7 / D30 / D90) rather than same-day — the realistic lens for subscription, gaming, and commerce apps with delayed revenue.
CAC payback: The time for a cohort’s contribution margin to recover its fully-loaded customer-acquisition cost — the finance-approved read a board runs on an app-growth motion.

Gaming + in-game advertising

GAP model: The In-game + Around-the-game + Away-from-the-game frame for gaming inventory — three buyer problems with different formats, KPIs, owners, and proof. Lets brands reallocate from a media-mix line, not a niche test budget.
Intrinsic in-game advertising: Ads rendered natively inside the game environment (virtual billboards, branded objects, in-scene video) — distinct from interstitials or rewarded units that interrupt play. Measured under the IAB/MRC Intrinsic In-Game (IIG) Guidelines.
In-game viewability: Whether an intrinsic in-game ad had the opportunity to be seen. The IAB/MRC IIG Guidelines v2.0 (2022) set a ≥1.5% screen-coverage threshold plus inactivity and fraud rules, adapting viewability logic to 3D environments.
Gaming brand safety: The controls that let a buyer place budget in games without content, fraud, or audience risk: suitability taxonomy, viewability, IVT controls, age-gating, verification, and creative review. The category’s biggest agency stall — and its wedge.
Gaming suitability: Matching a brand to appropriate game titles by genre, rating, content context, violence level, UGC / chat exposure, and category exclusions — the title-level discipline beneath "brand safety".
Rewarded video: An opt-in video format where players receive an in-game reward for watching. High completion rates, but measurement must account for incentive bias.
Playable ads: Interactive ad units that let a user try a mini-experience before acting (install, visit) — common in mobile-gaming user acquisition.
Console programmatic: Programmatic buying inside console games/environments. As of 2026 it remains emerging, not an at-scale buyable channel — Microsoft’s 2022 intrinsic in-game ad plan did not ship (its real ad move is a preroll tier on ad-supported Xbox Cloud Gaming, in testing); Sony’s remains exploratory. Validate supply before claiming it.
Gaming attention: Attention-based measurement (eyes-on, dwell) applied to gaming. The IAB/MRC Attention Measurement Guidelines (2025) and in-game attention work are standardizing it; no vendor owns a mature in-game attention product yet.
Esports sponsorship: Brand investment in competitive-gaming teams, leagues, tournaments, and broadcasts. High-value reach, but comparability and repeatability are harder than standard media — an Around-the-game surface.
Creator gaming: Marketing through gaming creators and streamers on Twitch, YouTube, and Discord — sponsorships, integrations, and community activations, measured via creator analytics, brand lift, promo codes, and MMM.
Gaming measurement: The stack that makes gaming legible to a media plan: exposure → verification → attribution → incrementality → MMM → decision. Explicit about what is counted, inferred, and modeled.
Media-mix reallocation: Moving gaming from a one-off innovation / test budget into a planned, repeatable media-mix line — the core commercial outcome this playbook is built to produce.
Brand lift: Survey-based measurement of shifts in awareness, consideration, or favorability caused by exposure — a primary outcome metric for upper-funnel in-game and around-the-game media.
Incrementality: The net-new effect of spend, isolated via geo-lift, holdouts, conversion lift, or matched-market / PSA tests — distinct from attribution, which allocates credit to conversions that may have happened anyway. Used in: BI / MMM / Decision Intelligence Performance & Native
CTV gaming: The convergence of connected-TV and gaming audiences and inventory — gaming-context CTV, cloud-gaming surfaces, and gaming-audience segments — extending reach with a measurement bridge back to outcomes.
Seller-defined gaming audiences: Audience segments defined by the seller / publisher (e.g. via the IAB Seller-Defined Audiences framework) to package gaming inventory by interest or behavior without third-party identifiers — an Away-from-the-game audience strategy.

Native + performance GTM

M-Sales: Marketing pulled into the revenue quota — category POV, account triggers, proof, field enablement, content, and event strategy built to create pipeline and accelerate deals, not to run brand campaigns.
D-Sales: Data and product marketing tied to the revenue quota — segmentation, use-case packaging, product telemetry, revenue analytics, and proof loops designed to improve sales velocity and ACV.
Native advertising: Ad formats that match the look, feel, and function of the surrounding content — recommendation widgets, in-feed units, sponsored content, native video, and commerce native. In 2026 the frame is outcome infrastructure, not cheap traffic.
Recommendation platform: A platform that ranks and serves content, product, or offer recommendations — the engine layer beneath native discovery, increasingly exposed via APIs and agent-readable signals.
Content discovery: Surfacing relevant content, products, or offers to a user in-feed or post-article — the demand-generation job native does before branded search or marketplace purchase.
Commerce media: Advertising tied to commerce intent and outcomes — onsite and offsite retail media, marketplace, and content-to-commerce — measured by sales, basket, and new-to-brand, not clicks. Used in: Retail / Commerce Media Measurement
Retail media incrementality: Whether retail-media-driven sales were incremental (net-new) or simply attributed to purchases that would have happened anyway — increasingly the question RMN buyers ask first. The IAB/MRC Retail Media Measurement Guidelines (January 2024) frame incrementality as the potential causal impact of marketing, measured via randomized controlled trials, synthetic controls, matched-market tests, or models. Used in: Retail / Commerce Media Measurement
ACV model: Annual contract value modeling — the per-deal economics ($144K–$230K+ in this playbook) that turn a pipeline into booked revenue and set rep-productivity expectations.
Rep productivity model: Booked revenue per FTE modeled against ACV, ramp, close rate, and account motion ($3.25M–$4.7M / FTE at maturity here) — operator math, not pitch math.
Sales velocity: How fast pipeline converts to revenue — a function of opportunity count, win rate, ACV, and cycle time; the metric M-Sales and D-Sales are built to improve.
Holdout test: An incrementality method that withholds ads from a matched audience or geo to measure the net-new effect of spend against a control.
MMM reconciliation: Reconciling a channel’s reported performance against a marketing-mix model (response curves, saturation, contribution) so budget decisions are cross-channel comparable — open-source options include Google Meridian and Meta Robyn.
Agentic buying: Goal-based, AI-assisted media buying where agents plan, activate, optimize, and measure campaigns — e.g. Trade Desk Kokai, Taboola Realize+, Reddit Max Campaigns (beta), and emerging agent-to-agent protocols (AdCP).
Quality-adjusted outcome: A performance metric that discounts for supply quality (viewability, IVT, MFA, brand safety, attention) so outcomes are comparable like-for-like, not gamed by cheap impressions.
Supply quality: The premium-ness and integrity of the inventory a platform sells — placement context, viewability, fraud controls, and brand safety — the answer to native’s clickbait / arbitrage legacy.
MFA (made-for-advertising): Low-quality sites built primarily to carry ads. Direct MFA spend has fallen sharply (ANA programmatic-transparency benchmarks), but it remains a buyer-quality concern enterprise platforms must address.
Premium native: Native inventory on quality publishers with transparent economics, viewability, and brand safety — the positioning that moves native from DR budgets toward brand and enterprise budgets.
Outcome infrastructure: The reframe at the heart of this playbook: native / recommendation as a measurable commerce-and-outcome layer a CMO, finance, and holdco can plan around — not a format or a traffic source.

Agentic transformation + commercial productization

Agentic transformation: The operating model for moving from scattered AI pilots to governed agentic workflows — decision rights, tool access, evaluation, fallbacks, semantic infrastructure, and measurable business outcomes.
Agentic workflow: A workflow where an AI agent plans and acts across signals, context, tools, and decisions under explicit governance — not a single prompt, but an end-to-end task with an owner and an outcome metric.
Agentic operating system: A six-layer model for designing agentic work across signals, context, tools, decisions, governance, and outcomes — so agents are built as a system, not as one-off automations.
Human-in-the-loop: A control pattern where a human reviews, approves, or overrides an agent before a consequential action is taken. The default for decisions an agent is not yet trusted to make alone.
Tool permissions: The allow-listed set of tools, APIs, and data an agent may read, write, or call, with explicit scopes. Defines the operational boundary of what an agent can do.
Output policy: The rules governing what an agent or collaboration environment is allowed to emit — formats, thresholds, redaction, and approvals. Distinct from access: a system can be allowed to read more than it is allowed to output. Used in: Data Clean Rooms, PETs & PAIR Multi-cloud orchestration Amazon Marketing Cloud
Fallback logic: The defined behavior when an agent is uncertain, fails, or hits a boundary — retry, escalate, hand off to a human, or stop. Without it, an agent fails silently or unsafely.
Decision control plane: The cross-workflow layer that governs how agentic decisions are made — decision rights, permissions, evaluation, monitoring, and fallback — so agents stay accountable as they scale.
Workflow owner: The named human accountable for a specific agentic workflow — its scope, evaluation, governance, and outcomes. Agentic work without an owner drifts and cannot be improved.
Outcome-level evaluation: Judging an agent on the business outcome it produces, not just task-level correctness — did the workflow move the metric it exists to move. The standard for deciding whether an agentic deployment is working.
Commercial productization: The operating model for turning a capability into a packaged product — offer architecture, pricing, proof, POC-to-production, and the repeatable enterprise revenue motion around it.
Offer architecture: The structure of a productized offer — buyer problem, what is in and out of scope, delivery model, and expansion path — that turns a capability into something defined and sellable.
Value metric: The unit a buyer grows on that price is tied to (seats, events, workflows, outcomes) so price scales with value delivered. The pivot of any durable pricing model.
POC-to-production: The path that converts a successful proof-of-concept into a renewed, productized, revenue-generating engagement — success criteria, conversion terms, commercial owner, and production scope.
Proof taxonomy: A classification of the proof each buyer needs — pilots, references, benchmarks, security, ROI — mapped to where it comes from, so a deal is not stalled on a missing receipt.
Buyer committee: The set of roles that must collectively approve an enterprise purchase — economic buyer, technical buyer, user, and blocker — each needing a tailored narrative and proof.
Marketplace listing: A public, self-serve discovery and distribution surface for a data product, app, or model (e.g. a cloud marketplace). Reduces distribution friction; not a GTM strategy on its own.
Private offer: A named-buyer, custom-terms distribution of a product — often via a cloud marketplace private offer. Higher margin and slower cycle than a public listing.
Native app: Application logic packaged to run next to a customer's data so the provider's IP and code stay protected — a productization option beyond a raw data share.
API product: A capability packaged and sold as a programmatic interface, with versioning, docs, rate limits, and pricing — productization for buyers who integrate rather than buy a UI.
Data product: A governed, packaged data asset sold or shared with defined contents, refresh cadence, quality guarantees, and access terms — the canonical thing being productized from a data stack.
Managed workflow: A productized offer where the vendor operates a recurring workflow as a service — packaged scope, SLA, and pricing — rather than handing over raw tooling.
Bespoke vs productized: The decision of which work to keep custom and which to package as a repeatable product — weighed on repeatability, margin, demand pattern, and delivery cost.

Engagements

Advisory Retainer: Ongoing 3, 6, or 12-month senior operator partnership for Series B Scaleup companies (75–200 employees · $10–30M ARR). Weekly cadence + crisis access.
GTM & BD Sprint: 6–8 week embedded execution sprint to convert audit or strategy into pipeline, partners, and decisions.
Market Entry Audit: 2–3 week diagnostic across 7 lenses, ships 10 artifacts including a 90-day GTM plan. Default first engagement.
Sponsored Market Intelligence Sprint: A disclosed, sponsor-funded advisory sprint. An independent advisor educates a defined market segment and synthesizes aggregated, anonymized intelligence for the sponsor. Confidential, independent, and never a lead list. Used in: Market Entry Audit Research & Measurement Science Services hub
Market readiness: How prepared a market segment is to understand, evaluate, and adopt a category — across understanding, urgency, and capability. Mapped in aggregate, never as individual scores.
Objection intelligence: The real, recurring reasons buyers hesitate in a category, captured in their own framing and aggregated — so positioning and education can address them honestly.
Category education: Vendor-neutral education that helps a market evaluate a category on its merits. The honest alternative to lead-capture “research,” and the core output a sponsored sprint funds.

Vocabulary vs diagnosis

Vocabulary is useful. Diagnosis is better.

If these terms are showing up inside a live company problem, the readiness scorecard will route you to the right playbook.

Take the readiness scorecard Explore common problems