We wrote this glossary to help you feel comfortable with the complex language in this ever-accelerating industry
Marketing attribution is an advanced discipline that helps marketers see greater returns on investment.
Visual IQ’s marketing attribution glossary provides definitions of common terms used in attribution measurement. Some of them may be new to you, and we want to demystify words you’ve heard but haven’t had a chance to clarify.
You’ll find definitions for some of the most important terms, including current vocabulary associated with data collection, cross channel marketing measurement and optimization, predictive analytics, and other related concepts.
Ad Server-Based Data Collection: The process of collecting digital user activity through a brand’s first-party or third-party ad server.
Addressable Channels: Channels where individual, user-level data (such as cookie data) is available to inform media spending and allocation decisions.
Advanced Attribution: A marketing methodology that leverages a multi-dimensional, algorithmic modeling approach to scientifically calculate and then fractionally assign conversion or brand engagement credit to every touchpoint and touchpoint attribute (ad size, placement, publisher, creative, offer, etc.) experienced by every converter and non-converter across all channels. It consists of three core capabilities: 1) measurement, 2) advanced measurement, and 3) optimization.
Algorithm: A mathematical formula that an attribution solution uses to assign conversion credit across the marketing channels and tactics to which a prospect was exposed prior to converting, as well as to predict the outcome of future marketing spend allocations.
Audience Attribution: A form of attribution that overlays demographic and behavioral data with media, response and customer data to uncover the marketing strategies and tactics that produce the greatest performance by specific audience segments. The output of audience attribution includes clear breakdowns of the customers with the highest propensity to convert, as well as the optimal combination of media and tactics needed to produce the best return for each audience segment. Audience segments may be defined by a marketer’s own first-party data, a third-party audience data aggregator, or a combination of the two.
Baseline (aka Brand Equity): The intrinsic value and level of performance against a set of key performance indicators (KPIs) derived solely from a brand’s recognition in the marketplace. It provides an initial value from which to evaluate the impact of incremental marketing investments.
Bottom-Up Modeling: A mathematical methodology that leverages individual, user-level data from addressable channels like online display and paid search to measure past marketing performance and predict future performance at the most granular level of data, such as creative, offer, keyword and more.
Brand Engagement: The measurement of the extent to which a consumer has a meaningful interaction with a brand (visits a landing page, views a video, downloads content, etc.) when exposed to a brand marketer’s middle- and upper-funnel media.
Channel: A digital and/or offline media category. Channels can be classified by paid, owned and earned, as well as addressable versus non-addressable.
Constraints: Factors associated with specific channels or tactics that limit the extent to which the media spend invested in them can be changed during the optimization process, and then the specific range of values in which media spend can vary to ensure the recommendations produced by an attribution solution can realistically be put in market given those factors. For example, the available inventory of branded search terms in a paid search marketplace is a constraint that limits the ability for increase spend on those terms more than 60% and decrease spend on those terms more than 100%.
Conversion Window: The period of time during which conversions are being tracked.
Container Tag: A delivery mechanism for tags that eliminates the need to place multiple data tracking codes on a website.
Cookie: A unique piece of code that is placed on a user’s browser by a tracking technology the first time that user accesses a given digital asset. Each piece of information a cookie tracks is called an “event,” and events are transmitted to ad servers or data collectors via codes called tags. For example, each time a user searches for an item online, it is recorded in their browser and later retrieved by ad servers to remember their behavior and better customize their search queries.
Cost: The cost that is paid for media.
Cost Per Thousand (CPM): The standard pricing model for many media channels (e.g. online display, video, etc.). Alternative pricing models are flat rate, pay per click (PPC) or cost per acquisition (CPA).
Cost Reconciliation: The process of associating cost data with paid channels and publishers to accurately calculate efficiency metrics like cost per acquisition.
Cross-Device Attribution: The process of applying algorithmic attribution to understand media impact across different devices beyond desktop computers, so that marketing spend can be optimized across smartphones, desktop, tablets and other devices.
Cross-Device Mapping: The process of matching a single user to two or more connected device ID’s associated with that user. This process may also be referred to as “cross-device matching,” “cross-device pairing,” or “cross-device bridging.” There are three different approaches to this process that can be used individually or together to maximize accuracy and scale: first-party deterministic, third-party deterministic, or third-party probabilistic.
Device ID: A unique identifier that can be used to identify a mobile device.
Device Type: A type of connected device technology that assigns its own unique device ID. Examples include: smartphones, tablets, desktops, and connected TVs.
Dimension (aka Attribute): A characteristic or feature associated with a marketing touchpoint. Each touchpoint may include over a dozen dimensions. For example, an online display ad may include the campaign, placement, publisher, ad size, creative, line of business, etc.
Earned Media: The free, publicity generated media produced by a brand’s fans. Examples include: Facebook likes, retweets, online reviews, word of mouth, etc.
Engagement Score: A KPI metric used to measure and optimize brand marketing campaigns. The score is typically a compilation of a set of events known as brand engagement activities, such as: first time website visits, rich-media ad interactions, video completions, etc.
Engagement Stack: A chronologically ordered list of all the marketing touchpoints experienced by an individual user.
Exogenous Factors: Factors external to the dimensions associated with marketing, including: seasonal factors (weather, holidays, etc.), economic factors (interest rates, gas prices, etc.) and competitive activities (changes to media tactics, new product launches, etc.) that lie outside of marketers' immediate control and may impact marketing effectiveness.
First Click/First Touch: A marketing attribution methodology that gives 100% of the conversion or brand engagement activity credit to the first marketing touchpoint experienced by a user prior to converting/engaging.
First-Party Data: Data that is owned by a brand and used to inform marketing. This data is typically collected from direct contact with the company’s customers.
First-Party Deterministic: One of three cross-device mapping approaches used to enable cross-device attribution. It may also be referred to as “brand authentication” since this approach uses the brand’s own first-party, authenticated user data. User authentication may include a website login, app interaction, email interaction, etc.
Halo (aka Halo Effect): A value that quantifies the degree to which different channels lift, or in some cases, cannibalize each other. For example, as the result of a halo effect, an advertiser may see an increase in branded search queries or conversions while running an online display campaign.
Impressions: In digital media, impressions are a measure of the number of times an ad is served. In offline media, impressions are a measure of the number of times an ad may have been seen.
Incrementality: A desired outcome (revenue, sales, leads, brand engagement, etc.) gained from a marketing activity that would not have been generated without that marketing activity.
Inter-Channel: The affinity between tactics used within one channel and those used within another, such as which online display ads drive searches on which keywords.
Intra-Channel: The affinity between different tactics used within the same channel, such as which non-branded keywords drive searches on which branded keywords.
Key Performance Indicator (KPI): The metric that a marketer uses to judge the success of a marketing initiative.
Last Click/Last Touch: An attribution methodology that gives 100% of the conversion credit to the last marketing touchpoint experienced by a user prior to converting/engaging.
Lookback Window: The defined period of time for which an ad can be expected to influence a user to convert or engage with a brand after exposure.
Match Rate: The number of user ID’s associated with a marketer’s touchpoint data that also match a set of user ID’s associated with another data provider (such as a cross-device data provider), divided by the total number of user ID’s associated with the marketer’s touchpoint data. Match rate is expressed as a percentage.
Media Attribute: The specific description of a given dimension (e.g. for the display dimension “ad size”, the media attribute might be “300x250”).
Model Validation: The process of verifying the accuracy of attribution measurement and the media spend recommendations it delivers by comparing the predicted results with the actual results.
Non-Addressable Channels: Includes channels like broadcast TV, radio, print, out-of-home, in-store displays, etc., where marketing messages are delivered to individuals who cannot be identified at a user-level.
Owned Media: All the communication assets that a marketer creates and has control over without having to make per-unit investments in order to expose them to the marketplace. Examples include a brand’s website, blog, mobile website, etc.
Paid Media: All of the advertising assets for which a marketer pays in order to expose them to the marketplace. Examples include TV commercials, print ads, online display ads, paid search, retargeting, etc.
Personally Identifiable Information (PII): Any data that can be used to identify a specific individual. Most marketers prefer to use non-PII data in order to protect consumer privacy.
Pixel: A transparent gif file or clear image placed within websites, ads and emails to track digital user activity (e.g. impressions, website visits, etc.) A pixel may also be referred to as a tag.
Pixel-Based Data Collection: The process of collecting digital user activity using through pixels.
Publisher: A media vendor that owns or manages media inventory, such as ad space, made available for purchase by marketers.
Point of Diminishing Returns: The point at which increased investment in a particular piece of media results in a decrease in the overall return (assuming all variables remain fixed). For example, the point of diminishing returns for a PPC campaign occurs when increasing the budget results in a decline in conversion rates and an increase in cost per acquisition.
Refresh: The process of building and applying an algorithmic attribution model within a defined conversion window and lookback window.
Refresh Rate (aka Refresh Cadence): The frequency with which an algorithmic attribution model is rebuilt for measurement and optimization purposes. Examples include daily, weekly, bi-weekly, etc.
Response Channels: Any channel that enables a customer or prospect to initiate a desired action in response to exposure to a marketing stimulation. Some advertisers call these “sales channels” or “revenue channels.” Examples include an eCommerce website, mobile website, traditional retail store, call center and more.
Rules-Based: An attribution methodology that distributes conversion credit across more than one marketing touchpoint using a manually or arbitrarily assigned weight. Examples of rules-based attribution models include even weighting, time decay, opener/advancer/closer, and linear methods.
Scenario Planning (aka “What-If?” Scenarios Planning): The use of predictive analytics to forecast future performance and produce customized media plans containing the optimal mix of channels and tactics needed to maximize marketing return while also accounting for constraints.
Segment: An identifiable group of individuals who share similar characteristics, needs or behaviors and who generally respond in a predictable matter to a given marketing stimulation.
Stimulation Channels: Any channel whose assets produce a marketing impression with a customer or prospect. Some marketers call these “impression channels,” “communication channels,” or “marketing channels.” Examples include paid search, online display, TV, radio, out of home, email, direct mail and more.
Tag: A piece of code placed on a website or an ad by which user data is collected. A tag may also be referred to as a pixel.
Taxonomy: Predefined classifications and naming conventions of touchpoint attributes, KPIs and other brand-specific categorizations. Taxonomies differ across organizations and ensure that the analyses, insights and recommendations derived from attribution map directly to a company’s unique jargon and business goals.
Third-Party Data: Data that is acquired by a marketer from an external source – such as a Data Management Platform (DMP) – for the purposes of marketing.
Third-Party Deterministic: One of three cross-device mapping approaches used to enable cross-device attribution. It may also be referred to as “third-party authentication” because this approach uses a third-party common login across devices (e.g. Google, Facebook, Twitter, etc.).
Third-Party Probabilistic: One of three cross-device mapping approaches used to enable cross-device attribution. It may also be referred to as “third-party inferred match” because this approach requires partnering with a cross-device technology vendor to make an inferred match across multiple devices to a single user. The connection is made using consumer behavior, relationship-based patterns, Wi-Fi linking, geo-clustering, as well as other variables.
Top-Down Modeling: A mathematical methodology that leverages summary-level data from both non-addressable channels and addressable channels to infer the relationships between different channels and tactics and deliver recommendations for optimization. The summary-level data that feeds a top-down model may include counts of individuals who were exposed to and/or took action upon various marketing initiatives; the particular date, time or location from which an advertisement was viewed; as well as exogenous factors, such as economic, seasonal and competitive data that have an impact on performance, such as interest rates, the weather, new product launches, etc.
Touchpoint: Any media interaction to which a customer or prospect is exposed. Touchpoints can include a wide range of interactions, from seeing a television commercial to conducting online price comparisons on a comparison shopping engine site, to clicking on a display ad or search result.
Trafficking: The process that places advertisements within media inventory.
Tribal Knowledge Factors: Any internal knowledge that lies within a company and may impact marketing effectiveness. Examples may include mergers and acquisitions, pricing changes, product launches, etc.
Unique Identifier/Unique ID: The mapping of multiple, disparate, anonymous user IDs associated with a unique user to a single, anonymous user ID that identifies the unique user across platforms, channels and devices.
Yield Curve: A line that plots the maximum number of theoretical “what-if?” media spend scenarios available to a marketer based on historical marketing performance data for purposes of forecasting marketing performance.