Glossary
Definitions of advertising technology terms and concepts used throughout OpenAdServe documentation.
Ad Tech Fundamentals
Ad Server
A technology platform that stores, manages, and serves advertisements to users. Ad servers make real-time decisions about which ads to display based on targeting criteria, budget constraints, and performance optimization. OpenAdServe is a publisher-first ad server designed for direct ad sales and header bidding integration.
Example usage: Publishers deploy an ad server to manage their direct-sold advertising campaigns and maximize revenue from their inventory.
See also: Campaign, Architecture Overview
Campaign
A top-level advertising initiative with a defined objective, budget, and timeline. Campaigns contain one or more line items that specify targeting and delivery settings. In OpenAdServe, campaigns represent the advertiser's overall advertising effort.
Example usage: A brand creates a Q1 awareness campaign with a $50,000 budget running across multiple placements.
See also: Line Item, Flight Dates
Click
A user interaction event when someone clicks on an advertisement. Clicks are tracked via signed URLs generated by the ad server and recorded in the analytics system. Click events trigger budget consumption for CPC campaigns.
Example usage: When a user clicks an ad, the click tracking URL verifies the token and records the event to ClickHouse.
See also: CPC, Impression, Token-Based Tracking
CPC (Cost Per Click)
A pricing model where advertisers pay only when users click their advertisements. CPC campaigns are converted to eCPM using estimated CTR for comparison with CPM campaigns during ad selection. This model reduces advertiser risk by charging only for engagement.
Example usage: An advertiser sets a $2.50 CPC bid, paying $2.50 each time someone clicks their ad.
CPM (Cost Per Mille)
A pricing model where advertisers pay per thousand impressions (mille is Latin for thousand). CPM is the most common pricing model for brand awareness campaigns. OpenAdServe uses CPM as the baseline for comparing different budget types.
Example usage: An advertiser sets a $10 CPM bid, paying $10 for every 1,000 impressions delivered.
See also: CPC, eCPM, Impression
Creative
The actual advertising content shown to users, including images, text, HTML, video, or other media. Each line item references one or more creatives that are served when the line item wins the ad selection process. Creatives contain the visual and interactive elements of the advertisement.
Example usage: A line item has three creative variants (different headlines) that rotate randomly to test messaging.
See also: Line Item, Placement
CTR (Click-Through Rate)
A performance metric calculated as (clicks / impressions) × 100. CTR measures how often users click an ad after seeing it. OpenAdServe uses estimated CTR to convert CPC bids to eCPM for fair competition with CPM campaigns.
Example usage: An ad with 5,000 impressions and 50 clicks has a 1% CTR.
See also: Click, Impression, CPC, CTR Optimization
Conversion
A valuable user action beyond clicking an ad, such as making a purchase, signing up, or downloading content. Conversions are tracked via custom events in OpenAdServe's analytics system. Conversion tracking enables advertisers to measure campaign ROI.
Example usage: A publisher tracks conversions when users complete a newsletter signup after clicking an ad.
eCPM (Effective CPM)
A normalized metric that converts different pricing models (CPM, CPC) into a comparable cost-per-thousand-impressions value. For CPC campaigns, eCPM = CPC × estimated CTR × 1,000. This enables fair comparison and ranking of campaigns with different budget types.
Example usage: A $2.50 CPC campaign with 1.5% estimated CTR has an eCPM of $37.50, allowing direct comparison with CPM campaigns.
Fill Rate
The percentage of ad requests that successfully return an advertisement. Calculated as (filled requests / total requests) × 100. A high fill rate indicates good inventory utilization and revenue potential. Publishers optimize fill rate through effective campaign management and header bidding.
Example usage: A placement with 10,000 requests and 9,500 filled has a 95% fill rate.
See also: Inventory, Ad Request
Flight Dates
The start and end dates during which a campaign or line item is eligible to serve. Flight dates define the temporal boundaries of advertising delivery. OpenAdServe only considers line items within their flight dates for ad selection.
Example usage: A seasonal campaign sets flight dates from November 1 to December 31 for holiday promotions.
See also: Campaign, Line Item, Pacing
Impression
A single instance of an advertisement being displayed to a user. Impressions are the fundamental unit of ad delivery and are tracked via impression pixels. CPM campaigns are charged based on impressions delivered.
Example usage: When a user's browser loads a page with an ad, an impression event is recorded via the tracking pixel.
See also: CPM, Click, Token-Based Tracking
Inventory
The total supply of available ad impressions a publisher can sell. Inventory represents monetization opportunities and is forecasted to help advertisers plan campaigns. OpenAdServe's forecasting engine predicts future inventory availability based on historical traffic patterns.
Example usage: A publisher queries their header placement inventory to see 500,000 available impressions next week.
See also: Forecasting, Fill Rate, Placement
Line Item
A specific advertising configuration within a campaign that defines targeting rules, budget allocation, pricing model, and delivery settings. Line items are the executable units of ad delivery. Each line item references creatives and competes for impressions based on priority and eCPM.
Example usage: A campaign contains three line items: one targeting mobile users, one for desktop, and one for tablet devices.
See also: Campaign, Creative, Targeting, Priority
Placement
A specific location on a publisher's website or app where advertisements can be displayed. Placements define the ad slot characteristics (size, format, context) and are the fundamental unit of inventory. Each ad request specifies which placement should be filled.
Example usage: A publisher defines a "header" placement for 728×90 leaderboard ads at the top of article pages.
See also: Inventory, Creative, Ad Request
Targeting & Delivery
Ad Request
An HTTP request from a publisher's website or app asking the ad server to select and return an appropriate advertisement. Ad requests include targeting data (device, location, context) and placement information. OpenAdServe processes requests through its ad selection pipeline.
Example usage: When a user visits a page, JavaScript makes an ad request to /ad with device type, GeoIP data, and placement ID.
See also: Placement, Targeting, Ad Selection
Forecasting
Predicting future inventory availability and campaign delivery based on historical traffic patterns and existing commitments. OpenAdServe's forecasting engine uses ClickHouse analytics to estimate available impressions, accounting for targeting, priority, and bid competition. Accurate forecasting prevents overselling and enables better campaign planning.
Example usage: An advertiser queries forecasted inventory to verify 100,000 impressions will be available for their targeting criteria next month.
See also: Inventory, Priority, Forecasting Documentation
Frequency Capping
Limiting the number of times a specific user sees the same advertisement within a defined time period. Frequency caps prevent ad fatigue and improve user experience. OpenAdServe tracks frequency caps in Redis using efficient token bucket counters.
Example usage: A line item sets a frequency cap of 3 impressions per user per day to avoid overexposure.
See also: Line Item, Impression, Pacing
Pacing
Distributing a campaign's budget evenly across its flight dates to prevent premature budget exhaustion. OpenAdServe uses dual-counter pacing that tracks both served impressions (for rate limiting) and recorded impressions (for accurate billing). This ensures stable delivery throughout the campaign period.
Example usage: A 30-day campaign with 300,000 impressions uses pacing to deliver approximately 10,000 impressions per day.
See also: Flight Dates, Dual-Counter Pacing, Pacing System
Priority
A numeric ranking that determines the order in which line items are considered during ad selection. Higher-priority line items can preempt lower-priority inventory. When line items have equal priority, they compete based on eCPM. Priority enables publishers to honor direct sales commitments while optimizing revenue.
Example usage: Guaranteed direct-sold campaigns run at priority 10, while programmatic demand runs at priority 5.
See also: Line Item, eCPM, Ad Selection
Targeting
Matching advertisements to specific audiences based on criteria like device type, operating system, geographic location, or custom dimensions. Targeting improves ad relevance and campaign performance. OpenAdServe supports device/OS/browser targeting, GeoIP, and custom key-value pairs.
Example usage: A line item targets iOS users in California with content_category=sports.
See also: Registered Dimensions, GeoIP, Line Item
Technical Terms
DSP (Demand-Side Platform)
A platform that enables advertisers and agencies to purchase ad inventory programmatically across multiple publishers. DSPs connect to ad exchanges and SSPs via RTB protocols. While OpenAdServe primarily serves publisher-direct campaigns, it can integrate with DSPs via Prebid Server.
Example usage: An advertiser uses a DSP to buy inventory across thousands of websites simultaneously.
See also: SSP, RTB, Programmatic
DMP (Data Management Platform)
A platform that collects, organizes, and activates audience data from various sources. DMPs enable sophisticated audience segmentation and targeting. Publishers can integrate DMP data with OpenAdServe via custom targeting dimensions.
Example usage: A publisher uses a DMP to create audience segments based on browsing behavior and passes segment IDs to OpenAdServe for targeting.
See also: Targeting, Registered Dimensions
GeoIP
Technology that determines a user's geographic location based on their IP address. OpenAdServe performs automatic GeoIP lookup for all ad requests, enabling location-based targeting without client-side data. GeoIP data includes country, region, and city.
Example usage: A restaurant chain targets users within 50 miles using GeoIP-derived coordinates.
See also: Targeting, Ad Request
Header Bidding
An advanced programmatic advertising technique where multiple demand sources bid on inventory simultaneously before the publisher's ad server makes a decision. Header bidding increases competition and revenue. OpenAdServe integrates with Prebid Server to support header bidding workflows.
Example usage: A publisher runs Prebid.js to collect bids from 8 SSPs, then passes the top bid to OpenAdServe for competition with direct campaigns.
See also: Prebid, Programmatic, SSP
OpenRTB
An open protocol for real-time bidding that standardizes communication between buyers (DSPs) and sellers (SSPs/exchanges). OpenRTB defines request/response formats for programmatic auctions. OpenAdServe can integrate with OpenRTB systems via Prebid Server.
Example usage: An SSP sends OpenRTB bid requests to multiple DSPs and returns the winning bid to the publisher.
See also: RTB, Programmatic, Header Bidding
Prebid
An open-source header bidding solution that enables publishers to offer inventory to multiple demand partners simultaneously. Prebid.js (client-side) and Prebid Server (server-side) collect bids and pass them to the ad server. OpenAdServe accepts Prebid bids for unified competition with direct campaigns.
Example usage: A publisher configures Prebid Server to send bid requests to 10 exchanges, then passes winning bids to OpenAdServe.
See also: Header Bidding, OpenRTB, Prebid Integration
Programmatic
Automated buying and selling of advertising inventory using technology platforms and real-time bidding. Programmatic advertising eliminates manual insertion orders and enables data-driven targeting at scale. OpenAdServe supports programmatic demand via Prebid integration while prioritizing direct publisher relationships.
Example usage: An advertiser uses programmatic buying to reach specific audiences across thousands of websites automatically.
See also: RTB, Header Bidding, DSP
RTB (Real-Time Bidding)
An auction mechanism where ad impressions are sold via instantaneous programmatic auctions. RTB enables advertisers to bid on individual impressions based on user data and context. Each auction completes in milliseconds during page load.
Example usage: When a user loads a page, an RTB auction determines which advertiser wins the impression based on bid amount and targeting.
See also: OpenRTB, Programmatic, Header Bidding
SSP (Supply-Side Platform)
A platform that helps publishers manage and sell their advertising inventory programmatically. SSPs connect publishers to multiple demand sources (DSPs, ad networks, exchanges) to maximize revenue. OpenAdServe functions as a publisher-first alternative to traditional SSPs for direct sales.
Example usage: A publisher uses an SSP to offer inventory to hundreds of demand partners simultaneously.
See also: DSP, Header Bidding, Programmatic
OpenAdServe Specific
AdCP (Ad Context Protocol)
An AI-powered automation protocol built on the Model Context Protocol (MCP) that enables natural language interaction with OpenAdServe. AdCP allows AI assistants to query inventory, create campaigns, forecast availability, and analyze performance through conversational commands.
Example usage: A user asks their AI assistant "Show me available inventory for publisher 1" and receives detailed forecasts without writing queries.
See also: AdCP Integration
Dual-Counter Pacing
OpenAdServe's advanced pacing system that tracks two separate impression counts: served impressions (for rate limiting) and recorded impressions (for accurate billing). This approach prevents race conditions and ensures fair budget consumption even when impression pixels fail to fire or are blocked.
Example usage: A line item tracks 10,000 served impressions for pacing purposes while 9,800 recorded impressions count toward budget consumption.
See also: Pacing, Pacing System
Registered Dimensions
A performance optimization feature that promotes frequently-used targeting attributes from generic key-value pairs to first-class indexed dimensions. Registered dimensions provide type validation, enum constraints, and optimized ClickHouse analytics queries. This feature improves targeting accuracy and reporting performance.
Example usage: A publisher registers content_category as an enum dimension with values [sports, news, tech, entertainment] for validated targeting and fast analytics.
See also: Targeting, Dimension Registration
Selector Interface
OpenAdServe's pluggable architecture for ad selection algorithms. The Selector interface allows custom ad decisioning logic to be implemented while reusing common filtering components. Publishers can implement custom selectors for specialized use cases (e.g., machine learning ranking, complex business rules).
Example usage: A publisher implements a custom selector that prioritizes ads based on predicted user lifetime value.
See also: Single-Pass Filter, Ad Selection
Single-Pass Filter
OpenAdServe's optimized ad selection algorithm that applies all filtering criteria (targeting, frequency caps, pacing, rate limits) in a single iteration through candidate line items. This approach delivers 6.3× better performance compared to traditional multi-stage filtering by minimizing memory allocations and iteration overhead.
Example usage: During ad selection, the single-pass filter evaluates 1,000 line items in under 2ms by checking all criteria simultaneously.
See also: Selector Interface, Ad Selection
Token-Based Tracking
A security mechanism that generates cryptographically signed URLs for impression and click tracking. Tokens prevent fraudulent event reporting by verifying that tracking URLs originated from legitimate ad responses. OpenAdServe validates tokens before recording events and updating budgets.
Example usage: An impression tracking URL includes a signed token that expires after 24 hours and can only be used once.
See also: Impression, Click, Token Service