Glossary of digital advertising terms-of-art, players, organizations, and technical processes, with an emphasis on those that impact good brand safety practice.
The use of software by a consumer to block digital advertising. The widespread adoption of ad blocking has been spurred on by poor ad experience practices, including intrusive ads that interrupt the browsing experience, overuse of tracking pixels to the point where website pages don’t load, and the infiltration of malware into the legitimate supply chain.
A digital marketplace where brands, agencies, publishers, and networks buy and sell display, video, and app ad inventory. An ad exchange allows buyers and sellers to use the same technical platform, deal parameters (ex. price, frequency), and data exchange services, improving campaign efficiency. Over time, the distinction between exchange and SSP has become less important, as many exchanges are now also SSPs, and many SSPs are now also exchanges. The terms are often now used interchangeably.
A consumer’s experience of the ads on a website’s page, including percentage of ads versus content, the size and format of each ad, and the functionality of each ad. Poor ad experiences contribute to the use of ad blockers by consumers. Some poor ad experiences can also trigger automatic blocking by some browsers - see the Coalition for Better Ads for best practices and ad formats to avoid.
Ad fraud can take many forms, including ad stacking, domain spoofing, botnets, and click fraud, all with the aim of siphoning marketing spend out from the legitimate supply chain. Brands and publishers are both hurt by marketing budgets being wasted on fake sites and traffic.
A method of inserting ads into an app, web page, etc., without the consent of the publisher or operator of that resource, often through browser plug-ins or apps that pose as legitimate utilities or services to trick a user into downloading them. These plug-ins and apps can house additional malware functionality dangerous to the user, such as keyloggers that can record account and password information, and they defraud publishers by suppressing the publisher’s sold advertising.
Attributes that describe the general makeup of the ad itself, including size, weight, CPU usage, duration, in-ad content, auto-play audio on/off, number of tracking pixels and server calls. Poor ad quality contributes to the use of ad blockers by consumers.
The fraudulent practice of placing multiple ads on top of each other in a single placement, with only the top ad being viewable by the user. In many cases, the advertiser pays for impressions on the unviewable ads, even if the end user never has the opportunity to see and interact with them.
Adjaceny to Unsafe Content/Misplacement
Advertising appearing next to or in front of content the marketer deems inappropriate for their brand. For example, an airline’s ads appearing on a news article about a plane crash.
Agency Trading Desk
An ad agency team and platform that specializes in buying media and audience programmatically. Trading desk platforms often sit on top of a demand side platform (DSP). Trading desks help brands improve campaign performance through data-driven optimizations.
A technique where one party will buy inventory cheaply and resell it at a higher price. It becomes a fraudulent practice if the buyer is unaware of the source of the inventory, or if the source or format are misrepresented (ex. Misrepresenting a display banner as video inventory).
Validating ads were served to the right audiences. For example: home buyers, geo targeting, etc.
Audience extension allows brands and agencies to target a premium publisher’s audience while they are browsing other sites across the web. Some publishers handle the buy directly, on behalf of the client, via their own demand side platform (DSP), while others prefer to send audience segments to the client via the client’s DSP or preferred exchange. Some publishers also make lookalike segments available, which allow a marketer to specify desired characteristics of a known audience segment, and target ads to a new set of users who share those characteristics.
A publisher practice of sending a new ad request in the same ad unit on a page at periodic intervals while the user remains on the page (ex. Every 30 seconds).
Preventing an ad from serving on a page that is deemed unsafe or undesirable, often with the use of keyword and semantic detection.
A list of domains or keywords that a brand is not willing to have its advertising appear on or near due to safety or suitability concerns.
Short for robot, this is a software program that carries out automated tasks on the Internet. There are good bots, such as those used for security scans, and bad bots, such as those used to view ads, watch videos, click on ads, etc. for the purposes of committing fraud.
The process of identifying bot traffic and impressions, and differentiating them from human traffic and impressions.
The process of detecting and filtering out bot traffic and impressions before inventory is bought or sold.
Nonhuman traffic designed by fraudsters to mimic users and inflate audience numbers
A group of computers or devices - often consumer devices - infected with malicious software that allows a fraudster to control device behavior for the purposes of committing ad fraud.
B2B companies hired by marketers or publishers to increase brand exposure with digital advertising.
To help make the web a better experience for users, most browsers pre-load some HTML, which can include ad requests, before the user accesses the content. In cases where the user does not eventually access the content, these pre-loaded ad requests are filtered out as invalid traffic.
Connected TV (CTV)
CTV includes any TV screen that can stream digital video via the internet, whether by itself (ex. smart TV) or through a separate over-the-top device such as Roku, AppleTV, FireStick, etc.
A method of stealing attribution credit, where a fraudster drops multiple cookies corresponding to multiple websites on a user’s browser, to make it appear that the user has visited those sites. This can be accomplished through popups, browser extensions, toolbars, or can be done when a user visits a fraudster’s site.
Data Management Platform (DMP)
A system for collecting first-party data, integrating first- and third-party data, and deploying this data in audience-focused marketing efforts. DMPs can help estimate reach for a given audience segment, connect data buyers and sellers in a clean, friendly environment, and help marketers and publishers understand the effectiveness of their data strategies.
The safeguarding of data from unauthorized access and use by third parties. For example: consumer data, publisher data, etc.
Traffic originating from datacenters, usually indicating automated bot activity rather than residential or corporate networks, which usually indicate human activity. This is typically considered invalid traffic, but is not necessarily malicious, as it can include legitimate uses, such as security scans. Additionally, proxy servers or other technologies may result in traffic appearing to originate from datacenters, while still being delivered to human users.
Demand Side Platform (DSP)
A platform that allows brands and agencies to connect with programmatic exchanges and publishers. DSPs handle the buy side of the real time bidding process, evaluating bid requests from exchanges and responding with a bid on the fly, based on the brand’s campaign parameters (ex. viewability thresholds, price limits, frequency targets, audience lists, etc.).
The practice of intentionally misrepresenting a site, device, etc. in the HTML or an ad request, with the intent to trick advertisers and ad exchanges into thinking the inventory is legitimate.
Any website or web page that is actively creating and/or distributing deliberately inaccurate content as news.
Frequency Cap Restricting (Capping)
The number of times a device can be served an ad from a given campaign in a specified period of time - ex. once per day. The desire is usually to limit the number of times an individual user will see a given campaign, but in practice, it is very difficult for ad servers and data management platforms (DMP) to connect a specific user to all of that user’s devices (ex. User X has a mobile phone, a laptop computer, and a Roku, but the campaign’s DMP has no way to see that these all belong to the same person), so frequency is often only possible on the device level (ex. User X gets the ad once per day on his mobile device, and once per day on his laptop, and once per day on his Roku).
General Invalid Traffic (GIVT)
Traffic that comes from known, nonhuman sources on publicly available IP lists. It can be identified through routine means of filtration by verification tools. Some GIVT is legitimate, such as datacenter traffic from routine site security scans, search engine crawlers, proxy servers, and pre-fetch or pre-rendered traffic. Some GIVT is malicious, such as bots and spiders or other crawlers masquerading as legitimate users, and hidden, stacked, covered, or otherwise unviewable ads.
A user’s device (browser, phone, or app) is modified without their authorization - usually by a malware infection - to request HTML or make ad requests that are not under their control and without their consent.
Interactive Advertising Bureau (IAB)
A global organization that develops industry standards, conducts research, and provides legal support for the online advertising industry. https://www.iab.com/
TThe ability to measure the age and gender of exposed audiences to ensure advertisers are reaching the desired target audience. This type of measurement is often validated by third party companies such as Nielsen.
Invalid Traffic (IVT)
Also referred to as Nonhuman Traffic (NHT) or Suspicious Activity Detection (SAD), it is online traffic generated from machines or other bot activity that interacts with digital ads. See also: GIVT and SIVT.
Inventory Quality Content
A set of attributes that differentiate types of content, enabling advertisers to make educated choices about the ad placements they are buying. Content classifications include categories such as "Unmoderated User Generated Content," "Breaking News," and "Mature Audiences." IQ guidelines are published by several trade groups, including the Trustworthy Accountability Group, the 4A’s, and the IAB.
A list of words a brand selects to either add to their ad targeting so that they will appear on pages using those words, or to add to a block list, to avoid appearing on pages using those words.
The consumer experience of a website or application loading or behaving slowly, often the result of poor ad creation or delivery. Latency is a major contributor to the use of ad blockers. See also: Ad Quality, Ad Experience.
Malware is an umbrella term for several types of malicious software – code specifically written to be harmful. It seeks to invade, damage, or disable computers, computer systems, networks, tablets, and mobile devices, often by taking partial control over a device's operations. Malware may include but is not limited to spyware, worms, bots, viruses, adware, phishing, auto-subscription, or unwanted changes to system configurations.
Media Rating Council (MRC)
A U.S.-based industry organization that aims to secure valid, reliable, and effective measurement services for the media industry through an audit and accreditation system. When MRC accredited, a verification provider’s measurement services meet the established criteria. http://mediaratingcouncil.org/
Activity that occurs before a programmatic purchase takes place. For example, a brand or agency may want to know the likelihood that inventory will meet certain criteria before bidding on it – that is, pre-bid.
Private Marketplace (PMP)
A programmatic auction that offers real-time-bidding in a closed, invitation-only auction. There are many flavors of PMPs. They can bet set up between demand side platforms (DSPs) and exchanges/supply side platforms (SSPs), between DSPs and publishers, between SSPs and publishers, and so on, and can be either one-on-one, or include multiple buyers and sellers. PMPs, when handled appropriately by responsible sell side partners, can offer assurance to buyers that they are not buying on fraudulent inventory, and can offer a "first look" opportunity into quality, sometimes exclusive or rare inventory that cannot be found in the open auction.
The automated buying and selling of digital media, as opposed to more traditional methods of transacting that rely heavily on manual data entry in multiple systems by multiple people, and tools external to campaign planning and execution systems, such as Excel. In programmatic advertising, buyers and sellers can take advantage of "pipes" connecting many platforms across the ecosystem to remove some of the inefficiencies of the process. For example, often, IOs are not needed, and reporting and billing are handled within the platform(s) rather than over email. The fact that there are thousands of platforms in use, some older and some newer, and that the connections do not always work as expected, means that there is still a large role for humans on both sides of a transaction to play in partnering to make deals work well. Because of this need for close collaboration in setup, monitoring, and troubleshooting, some practitioners on the front lines call it "programmanual."
Programmatic Direct is the striking of a deal between a brand or agency and publisher, rather than the buyer going through intermediaries. The deal could be for a guaranteed, fixed budget, or it could be for a non-guaranteed budget for an agreed upon bid floor. The key benefits of a buyer working directly with a publisher, as opposed to buying indirectly, is reduced "tech tax" - the amount of budget that goes towards paying various intermediaries - which puts more budget toward working media, and the knowledge that the inventory being purchased is clean and fraud-free.
Real-Time Bidding (RTB)
RTB is a very fast, nearly literally real time exchange between the buy and sell side within programmatic. When a publisher’s page loads and the ad requests are made, the publisher’s supply side platforms (SSPs) send bid requests through the exchanges to the demand side platforms (DSPs). The DSPs evaluate the bid requests, looking at historical performance of the inventory, available user matching information, and any relevant campaign parameters, and then decide to either pass or respond with a bid. The bid goes back to the sell side systems, and if the bid wins the auction, the DSP’s ad creative serves to the user. This process typically takes 200 milliseconds or less.
A list of pre-approved domains or keywords that a brand finds acceptable, or safe, for its advertising to appear on or near, regardless of other brand safety settings that may already be in place. Also known as an Allow List.
Sophisticated Invalid Traffic (SIVT)
Nonhuman traffic that is more difficult to detect than General Invalid Traffic (GIVT), and requires advanced analytics, multipoint corroboration/coordination, or significant human intervention to analyze and identify. Key examples include hijacked devices, hijacked tags, adware, malware, incentivized browsing, misappropriated content (if applicable), falsified viewable impression decisions, and cookie stuffing.
Supply-Side Platform (SSP)
An SSP is a sell-side platform that allows publishers to more easily make their inventory available to buyers through exchanges, networks, and demand side platforms (DSPs). Over time, the distinction between exchange and SSP has become less important, as many exchanges are now also SSPs, and many SSPs are now also exchanges. The terms are often now used interchangeably.
A 1x1 pixel-sized transparent image that provides information about an ad's placement, such as whether an ad has been served, the domain the ad served on, whether the ad was viewable, etc.
Trustworthy Accountability Group (TAG)
A cross-industry accountability program that works to create transparency in business relationships and transactions in the digital ad industry, while continuing to enable innovation. https://www.tagtoday.net/
Viewability is an online advertising metric that measures the number of impressions viewed by real, human users. A served impression does not necessarily count as a viewed impression, if it is served outside a live window or is served to a bot. The Media Rating Council publishes an official industry definition of what should be counted as a viewable ad, based on the percentage of the creative and the duration in seconds that it is in an active browser or screen.