RTB, DSP, CPM, CTV: The Ad Tech Glossary That Actually Makes Sense

Walking into a programmatic advertising conversation without knowing the terminology feels like attending a medical conference where everyone assumes you went to med school. The industry has evolved so rapidly over the past fifteen years that the vocabulary has become genuinely intimidating, with layer upon layer of acronyms describing platforms, pricing models, targeting methods, and measurement frameworks that all interconnect in complex ways. This glossary exists to demystify that vocabulary by explaining terms in context, with real examples that show how each piece fits into the larger ecosystem of digital advertising. Rather than providing sterile dictionary definitions, we will explore how these terms relate to one another and how understanding them helps you make better decisions about your advertising strategy.

The programmatic advertising industry generated over $140 billion in revenue in the United States alone in 2024 according to the Interactive Advertising Bureau, and that growth has been accompanied by an explosion of specialized terminology. Every platform, pricing model, and targeting method has its own acronym, and many of these terms are used interchangeably or incorrectly even by people who work in the industry. The confusion is compounded by the fact that technology evolves faster than language, so terms that meant one thing five years ago may have shifted in meaning as new capabilities emerged. For marketers trying to evaluate platforms or agencies, this vocabulary barrier creates real problems because you cannot assess what you are buying if you do not understand the words being used to describe it. This guide will walk through the major categories of programmatic advertising terminology, providing context and examples that make each term concrete rather than abstract.

When advertisers want to purchase digital ad inventory programmatically, they use a suite of interconnected platforms that automate targeting, bidding, and optimization. The most fundamental tool on the buying side is the demand-side platform, universally abbreviated as DSP. A DSP is software that allows advertisers to purchase ad impressions across multiple ad exchanges and publisher sites through a single interface, using real-time bidding to buy impressions that match specific audience criteria. For example, a restaurant chain might use a DSP to bid on impressions served to users within five miles of their locations who have searched for dining options in the past week. The DSP evaluates millions of available impressions per second, calculates the value of each impression based on the advertiser's targeting criteria and budget constraints, and submits bids automatically. Major DSPs include The Trade Desk, Google's Display & Video 360, Amazon DSP, and Verizon Media DSP, each offering different strengths in terms of inventory access, data integration, and reporting capabilities.

Supporting the DSP are data platforms that provide the audience intelligence needed to target effectively. A data management platform, or DMP, is a system that collects, organizes, and activates first-party data from your own properties alongside second-party and third-party data from external sources to build targetable audience segments. If you run a healthcare clinic, your DMP might combine website visitor behavior from your own site with demographic data from a third-party provider to create a segment of women aged 35 to 50 in your region who have visited pages about specific treatments. That segment is then pushed to your DSP to inform bidding decisions. DMPs were the dominant data solution for programmatic advertising throughout the 2010s, but they have faced challenges as third-party cookies have been deprecated and privacy regulations have tightened. In response, the industry has increasingly adopted customer data platforms, or CDPs, which focus primarily on first-party data and provide deeper, more persistent customer profiles that extend beyond advertising into email marketing, personalization, and customer service.

The distinction between DMPs and CDPs matters because it reflects a fundamental shift in how advertisers think about data. A DMP is designed for anonymized audience segments used in ad targeting, typically built on cookies and device IDs that provide shallow, transient identifiers. A CDP is designed for known customer profiles that persist over time and integrate with your CRM, allowing you to build unified customer records that include names, emails, purchase history, and cross-channel behavior. For example, Senova's platform functions as a CDP with built-in programmatic activation, meaning you can build audience segments from your lead and customer data and push them directly to managed ad campaigns without needing a separate DMP. This integration is particularly valuable for small and mid-sized businesses that do not have the resources to manage multiple disconnected data platforms.

Some larger advertisers work with a trading desk, which is a specialized team or agency that manages programmatic ad buying on behalf of the advertiser, often using proprietary technology and access to premium inventory. A trading desk might sit inside a large media agency or operate as an independent entity, and it typically offers services beyond basic DSP access, including custom audience research, creative production, and advanced optimization strategies. For example, a national retail brand might engage a trading desk to manage a $10 million programmatic budget across display, video, and connected TV, with the trading desk responsible for negotiating private marketplace deals, testing creative variations, and delivering detailed performance reports. Trading desks were more common in the early days of programmatic advertising when DSPs were less user-friendly, but they remain relevant for advertisers who want managed services without building in-house expertise.

On the opposite side of the transaction are the platforms that help publishers monetize their content by selling ad inventory to the highest bidder. The supply-side platform, or SSP, is software that allows publishers to offer their ad inventory to multiple ad exchanges and demand sources simultaneously, maximizing fill rates and revenue by creating competition among buyers. A publisher like an online news site might use an SSP to make their display and video ad slots available to dozens of DSPs at once, with the SSP automatically accepting the highest bid for each impression. Major SSPs include Google Ad Manager, Magnite, PubMatic, and OpenX, and most publishers work with multiple SSPs to diversify their demand sources and avoid dependence on a single platform.

Connecting buyers and sellers is the ad exchange, which is a digital marketplace where DSPs and SSPs meet to trade ad impressions in real-time auctions. An ad exchange operates like a stock exchange, with bids and asks being matched in milliseconds based on price and targeting criteria. Google's AdX is the largest ad exchange, handling billions of impressions daily across display, video, and mobile inventory. When you visit a website and see an ad load, that ad was likely purchased through an ad exchange auction that occurred in the 100 to 200 milliseconds between your browser requesting the page and the ad rendering on your screen. The exchange receives bid requests from the SSP, broadcasts those requests to connected DSPs, collects bids, determines the winner based on auction rules, and sends the winning ad creative back to the publisher's ad server for display.

Publishers can also sell inventory directly to specific buyers through private marketplaces, often abbreviated as PMP. A private marketplace is an invitation-only auction where publishers offer premium inventory to a curated group of buyers, often at higher floor prices than open exchanges. For example, a major sports media site might create a PMP for endemic advertisers in the sports and fitness category, guaranteeing them access to homepage placements during live game coverage in exchange for higher CPMs. PMPs offer publishers more control over who advertises on their site and allow buyers to secure access to high-value inventory that might not be available on open exchanges. Some publishers go further with programmatic guaranteed or programmatic direct deals, which are one-to-one agreements where a buyer commits to purchasing a specific volume of impressions at a fixed price, combining the automation of programmatic technology with the certainty of traditional direct sales.

Understanding the difference between open exchanges, private marketplaces, and programmatic guaranteed deals is crucial because it affects both pricing and inventory quality. Open exchanges offer the broadest reach and lowest prices but also the highest risk of ad fraud, brand safety issues, and low-quality placements. Private marketplaces provide better quality control and premium placements at higher prices. Programmatic guaranteed deals offer the highest quality and predictability but require upfront commitments and negotiations. A well-structured programmatic strategy typically uses a mix of all three, allocating budget based on campaign goals and risk tolerance.

Programmatic advertising supports multiple pricing models, each aligned with different campaign objectives and advertiser preferences. The most common model is cost per thousand impressions, universally abbreviated as CPM, where the advertiser pays a fixed amount for every one thousand times their ad is displayed, regardless of whether anyone clicks or converts. CPM pricing is ideal for brand awareness campaigns where the goal is to reach as many people as possible within a target audience. For example, a new restaurant opening in a city might run a CPM-based display campaign targeting local residents aged 25 to 45, paying $8 CPM to serve 500,000 impressions over two weeks. The total cost would be $4,000, and the restaurant would measure success based on how many people saw the ad rather than how many clicked.

Cost per click, or CPC, shifts the pricing model so that advertisers only pay when someone clicks on their ad, transferring some risk from the advertiser to the publisher or platform. CPC pricing is common for search ads and social media ads where click-through rates are relatively high and the advertiser's goal is to drive traffic to a landing page or website. For example, a law firm might run a Google search ad targeting the keyword "personal injury attorney" with a maximum CPC bid of $45, meaning they will pay up to $45 each time someone clicks their ad. CPC pricing aligns cost directly with engagement, making it attractive for performance-focused advertisers, but it can be expensive in competitive categories where clicks are scarce and valuable.

Cost per action or cost per acquisition, both abbreviated as CPA, takes performance pricing even further by charging advertisers only when a specific action is completed, such as a form submission, purchase, or app install. CPA pricing is ideal for direct response campaigns where the advertiser has a clear conversion goal and known customer lifetime value. For example, an e-commerce brand selling subscription meal kits might run a CPA campaign targeting health-conscious consumers, agreeing to pay $30 for every new subscription signup. The platform or publisher bears all the risk of delivering results, which typically means CPA campaigns require higher effective CPMs to compensate for the uncertainty. Many CPA deals are structured as rev-share or performance partnerships where the publisher or affiliate takes a percentage of each sale rather than a fixed fee.

Return on ad spend, or ROAS, is a measurement rather than a pricing model, but it is frequently used as the performance benchmark for campaigns priced on CPM or CPC. ROAS is calculated by dividing revenue generated by the campaign by the amount spent on the campaign, expressed as a ratio or percentage. For example, if you spend $5,000 on a programmatic campaign and it generates $20,000 in tracked sales, your ROAS is 4:1 or 400%. Many advertisers set target ROAS thresholds for their campaigns, such as "we need at least 3:1 ROAS to be profitable," and optimize their targeting and creative to hit those targets. Platforms like Google and Facebook offer automated bidding strategies that optimize toward target ROAS, adjusting bids in real-time to maximize revenue while staying within budget.

Closely related is cost per lead, or CPL, which is a CPA model specific to lead generation campaigns where the advertiser pays for each qualified lead delivered, such as a completed contact form or phone call. CPL pricing is common in industries like home services, healthcare, insurance, and financial services where the sales cycle is long and leads are valuable. For example, a roofing company might run a programmatic display campaign targeting homeowners in storm-affected areas, agreeing to pay $50 per qualified lead. The platform uses visitor identification technology to track which website visitors came from the campaign and which of those visitors submitted contact forms, attributing leads back to the campaign and billing accordingly.

Programmatic advertising derives much of its value from sophisticated targeting capabilities that allow advertisers to reach specific audiences based on behavior, demographics, location, and context. Contextual targeting places ads based on the content of the page where the ad appears, without using personal data or tracking individual users. For example, a sporting goods retailer might use contextual targeting to show ads on articles about marathon training, hiking gear reviews, and fitness tips, reaching people who are actively interested in athletic activities regardless of their individual browsing history. Contextual targeting has seen a resurgence as third-party cookies have been deprecated, and many advertisers now use advanced natural language processing tools to analyze page content and match ads to contextually relevant environments.

Behavioral targeting uses data about an individual's past online behavior, such as websites visited, searches performed, and content consumed, to predict their interests and intent. For example, someone who has visited multiple car dealership websites, read reviews of SUVs, and used a car loan calculator is likely in the market for a vehicle, making them a valuable target for automotive advertisers. Behavioral targeting typically relies on cookies, device IDs, or authenticated user IDs to track users across sites and build profiles that inform ad serving decisions. This approach is highly effective for performance campaigns but faces increasing limitations as browsers block third-party cookies and users opt out of tracking.

Geofencing is a location-based targeting method that triggers ads when a user enters a defined geographic area, such as a radius around a store, competitor location, or event venue. For example, a coffee shop might set up a geofence around nearby office buildings, serving ads to people who enter those areas during morning commute hours, offering a discount to entice them to visit. Geofencing relies on GPS, WiFi, and cell tower data from mobile devices to determine when a user enters or exits a target zone. More advanced versions of geofencing include address-level targeting, where ads are served to devices associated with specific street addresses, and dwell time targeting, which only triggers ads if the user spends a minimum amount of time in the geofenced area. Senova offers geofencing as part of its campaign activation platform, allowing businesses to target competitor locations, event venues, and custom geographic boundaries.

Lookalike audiences are algorithmically generated segments that resemble an advertiser's existing customers based on shared characteristics and behaviors. For example, a medical spa might upload a list of their 500 best customers to a DSP, which analyzes those customers' demographics, online behavior, and purchase patterns, then identifies 50,000 other users who exhibit similar traits. Lookalike modeling is one of the most powerful targeting methods available because it allows advertisers to expand beyond their known customer base while maintaining relevance. The effectiveness of lookalike audiences depends heavily on the quality and size of the seed audience, with larger and more homogeneous seed lists typically producing better results.

Retargeting, also called remarketing, targets users who have previously interacted with your brand, such as visiting your website, viewing a product page, or abandoning a shopping cart. For example, an online furniture retailer might serve display ads featuring the specific couch a user viewed but did not purchase, offering a 10% discount to encourage them to complete the transaction. Retargeting campaigns typically achieve much higher conversion rates than prospecting campaigns because they focus on users who have already demonstrated interest. Most retargeting relies on cookies or pixel-based tracking, where a snippet of code on your website drops a cookie in the visitor's browser, allowing you to recognize them later and serve targeted ads as they browse other sites.

Effective programmatic advertising requires robust measurement to understand what is working and optimize toward better results. Attribution modeling assigns credit for conversions to the various touchpoints a user encountered along their path to purchase, helping advertisers understand which channels and campaigns are driving results. Single-touch attribution models, such as first-click or last-click attribution, assign 100% of the credit to a single touchpoint, making them simple but often misleading. Multi-touch attribution models, such as linear, time-decay, or algorithmic attribution, distribute credit across multiple touchpoints based on their relative contribution to the conversion. For example, a user might see a display ad, click a search ad, and then convert after receiving an email, and a time-decay attribution model might assign 20% credit to the display ad, 30% to the search ad, and 50% to the email.

Viewability measures whether an ad had the opportunity to be seen by a user, defined by the Media Rating Council as at least 50% of the ad's pixels being visible on screen for at least one second for display ads or two seconds for video ads. Viewability is important because advertisers should not pay full price for ads that load below the fold and are never scrolled into view. According to industry benchmarks, desktop display ads typically achieve 60 to 70% viewability, while mobile display ads often struggle to exceed 50% viewability due to smaller screens and faster scrolling behavior. Many advertisers now buy on a viewable CPM basis, meaning they only pay for impressions that meet the viewability threshold, shifting the risk of non-viewable placements to publishers.

Brand lift studies measure the impact of advertising on brand perception metrics such as awareness, consideration, favorability, and purchase intent. A brand lift study typically involves splitting an audience into a control group that does not see your ads and an exposed group that does see your ads, then surveying both groups to measure differences in brand perception. For example, a consumer goods brand might run a video campaign on connected TV and conduct a brand lift study to determine that the campaign increased brand awareness by 12 percentage points and purchase intent by 8 percentage points among the exposed audience. Brand lift studies are particularly valuable for upper-funnel campaigns where direct conversion tracking is difficult or where the primary goal is to shift perception rather than drive immediate sales.

Conversion tracking measures the actions users take after seeing or clicking an ad, such as purchases, form submissions, phone calls, or app installs. Conversion tracking relies on tracking pixels, postback URLs, or server-to-server integrations that report conversions back to the advertising platform. For example, when someone completes a contact form on your website, a conversion tracking pixel fires and sends data back to your DSP indicating that the campaign drove a lead. That conversion data is then used to calculate metrics like cost per conversion, conversion rate, and ROAS, which inform optimization decisions. Accurate conversion tracking is essential for performance campaigns, and many advertisers struggle with attribution when conversions happen offline or across devices.

Frequency capping limits the number of times an individual user sees your ad within a specified time period, preventing ad fatigue and wasted impressions. For example, you might set a frequency cap of three impressions per user per day, ensuring that your budget is spread across a larger audience rather than bombarding the same users repeatedly. Research consistently shows that ad effectiveness declines sharply after a user has seen the same ad three to five times, and excessive frequency can actually harm brand perception by annoying potential customers. Most DSPs and ad servers offer frequency capping controls, and best practice is to set caps appropriate to your campaign length and creative rotation strategy.

Digital advertising spans a wide range of channels and formats, each with its own specialized terminology. Display advertising refers to banner ads, sidebar ads, and other visual ad units that appear on websites and apps, typically sold on a CPM basis. Display is the oldest and most widespread form of programmatic advertising, with standard sizes defined by the Interactive Advertising Bureau, including the 300x250 rectangle, 728x90 leaderboard, and 160x600 skyscraper. Display ads can be static images, animated GIFs, or HTML5 rich media units, and they are used for both brand awareness and direct response campaigns. Despite concerns about banner blindness and declining click-through rates, display advertising remains a core component of most programmatic strategies due to its broad reach and relatively low cost.

Native advertising is a format designed to match the look and feel of the surrounding content, appearing as sponsored articles, recommended content widgets, or in-feed social posts. Native ads typically achieve higher engagement rates than standard display ads because they blend into the user experience rather than interrupting it. For example, a travel company might run a native ad on a news site that appears as a recommended article titled "10 Hidden Beaches You Need to Visit This Summer," linking to a landing page with booking options. Native advertising platforms like Taboola and Outbrain specialize in distributing native ads across publisher networks, while many DSPs also offer native ad buying capabilities.

Video advertising includes pre-roll ads that play before online video content, mid-roll ads that play during content, and post-roll ads that play after content, as well as outstream video ads that play in article content or social feeds. Video ads are typically sold on a CPM basis, with premium video inventory commanding higher prices than display due to higher engagement and completion rates. According to the Interactive Advertising Bureau, digital video advertising accounted for over $50 billion in spending in 2024, with mobile video and connected TV driving the majority of growth. Video completion rate, or VCR, is a key metric for video campaigns, measuring the percentage of viewers who watch the entire ad, and most advertisers consider a VCR above 70% to be strong performance.

Connected TV, or CTV, refers to television sets that connect to the internet to stream content from services like Netflix, Hulu, Disney+, and YouTube, while over-the-top, or OTT, refers to the content delivery mechanism that bypasses traditional cable and satellite providers. CTV advertising allows brands to serve targeted ads during streaming content, combining the reach and impact of television with the targeting and measurement capabilities of digital advertising. For example, a luxury car brand might target households with incomes above $150,000 who have shown interest in automotive content, serving a 30-second ad during premium streaming content. CTV has become one of the fastest-growing channels in programmatic advertising, with advertisers increasingly shifting budgets from traditional TV to streaming platforms that offer better targeting and accountability.

Digital out-of-home, or DOOH, refers to digital billboards, transit ads, mall displays, and other out-of-home advertising formats that use digital screens instead of printed materials. DOOH inventory can be bought programmatically, with ads served dynamically based on factors like time of day, weather, traffic conditions, and audience demographics detected through mobile location data. For example, a quick-service restaurant might serve breakfast ads on highway billboards during morning rush hour and lunch specials during midday. DOOH combines the mass reach of traditional outdoor advertising with the flexibility and targeting of programmatic, and it is increasingly integrated with mobile campaigns that serve follow-up ads to users who passed by DOOH placements.

Audio advertising includes ads served during streaming music on platforms like Spotify and Pandora, podcast ads inserted dynamically into podcast episodes, and ads on streaming radio services. Audio ads are typically 15 to 30 seconds long and are sold on a CPM basis, with targeting based on listening behavior, demographics, and location. For example, a fitness brand might target users who listen to health and wellness podcasts, serving a 30-second ad promoting their new workout app. Podcast advertising has grown particularly rapidly, with many advertisers valuing the engaged, loyal audiences that podcasts attract. Dynamic ad insertion technology allows programmatic buyers to place ads into podcast episodes in real-time, enabling more sophisticated targeting and measurement than traditional baked-in podcast ads.

Understanding these terms individually is useful, but the real power comes from understanding how they interconnect to form a complete programmatic advertising ecosystem. When you launch a campaign, you typically start by defining your audience using data from a CDP or DMP, then activate that audience through a DSP that connects to multiple ad exchanges and SSPs to bid on relevant inventory. Your DSP uses real-time bidding to evaluate millions of available impressions per second, submitting bids for impressions that match your targeting criteria based on contextual signals, behavioral data, geofencing parameters, or lookalike modeling. When your bid wins, your ad is served on display, native, video, CTV, DOOH, or audio inventory, and you pay based on CPM, CPC, or CPA depending on your campaign structure.

As your campaign runs, measurement systems track viewability to ensure your ads are actually seen, frequency capping prevents overexposure, and conversion tracking measures the actions users take after seeing your ads. Attribution models assign credit to different touchpoints, helping you understand which channels and tactics are driving results, and brand lift studies measure changes in perception metrics that may not show up in immediate conversion data. You optimize your campaign based on metrics like ROAS, CPL, or video completion rate, adjusting your targeting, creative, and bidding strategy to improve performance over time. Publishers on the other side of the transaction use SSPs to maximize revenue from their inventory, offering premium placements through private marketplaces and programmatic guaranteed deals to buyers willing to pay higher prices for better quality and transparency.

This entire process happens in a fraction of a second, with technology automating decisions that would be impossible for humans to make manually. The terminology exists to describe each component and process within this ecosystem, and understanding it allows you to participate in the conversation, evaluate vendors and platforms, and make informed decisions about your advertising strategy. You do not need to be an expert in every acronym or technology to run effective programmatic campaigns, but you do need to understand the basic categories of tools, pricing models, targeting methods, and measurement frameworks so that you can ask the right questions and assess whether you are getting good results.

When you meet with a programmatic advertising vendor or agency, they will inevitably use many of these terms in their pitch. Your ability to understand and question that language directly impacts your ability to evaluate whether they are offering a genuine solution or obscuring weaknesses with jargon. If an agency tells you they use a proprietary trading desk with access to premium PMPs and guaranteed viewability above 80%, you now know to ask which SSPs they work with, what floor prices they are paying, and how they measure viewability. If a platform promises to deliver a 5:1 ROAS using lookalike audiences and multi-touch attribution, you can ask about their seed audience requirements, whether they use deterministic or probabilistic attribution models, and what conversion tracking implementation is required.

For businesses working with Senova, much of this complexity is abstracted away through managed services, but understanding the underlying terminology still helps you have more productive conversations about strategy and results. When your account manager explains that your campaign is running on a mix of open exchange and PMP inventory with a blended CPM of $12 and achieving a 3.2:1 ROAS through a combination of geofencing and behavioral targeting, you will understand what those metrics mean and whether they align with your goals. Senova's platform integrates visitor identification, audience intelligence, and campaign activation into a single workflow, meaning you do not need to manage separate DMPs, CDPs, and DSPs, but knowing what those components do helps you appreciate the value of that integration.

The terminology in this glossary represents the current state of programmatic advertising, but the industry continues to evolve rapidly, with new technologies and terms emerging regularly. Privacy regulations like GDPR and CCPA are forcing changes in how targeting and measurement work, with increased focus on first-party data, contextual targeting, and privacy-preserving measurement methods. The deprecation of third-party cookies is driving adoption of alternative identifiers like Unified ID 2.0 and probabilistic device graphs that attempt to maintain cross-site tracking without cookies. New channels like retail media networks and commerce media are creating new inventory sources and measurement opportunities. Staying current with terminology and trends is an ongoing process, but the foundational concepts covered here provide a solid framework for understanding how programmatic advertising works and how to apply it effectively in your business.