Lytics Glossary

This glossary defines terms used across the Lytics platform and documentation.

Your unique Lytics account identifier that is used to reference your account externally such as with the JavaScript Tag, APIs, etc. Find your account ID in the Lytics UI by navigating to the details section under the Manage Account menu.

Options to manage your Lytics account including access permissions, API tokens, and more. See Account Management for more.

A short integer at the end of your Lytics account URL. For example: https://activate.getlytics.com/profile?**aid=1234**

Technical instructions that detail the execution of a process, often used in reference to machine learning.

An option that allows Lytics audiences to be shared with integrations, such as Facebook. By default, Lytics audiences are NOT enabled for public API access so you will need to manually select this Audience Configuration option to allow Lytics audiences to be shared with client-side tags/ pixels, Pathfora, or the Lytics personalization API endpoints.

An API Token, also referred to as an “API Key”, is a unique identifier used to authenticate a user, developer, or call a program to an API.

A user field that contains any data associated with the user profile. Examples include audience membership, topic affinities, impressions on a particular campaign, etc.

A group of users, or collection of user profiles within Lytics.

Lytics' tool for creating specific groups of users that can be targeted onsite or in downstream tools with Lytics Experiences. Read more on Building Audiences.

Audience Discovery is a report powered by Lytics data science that provides a comparison of two audiences and helps you discover the behavioral attributes that influence different levels of customer engagement.

Actions taken by users on your various channels such as opening an email, clicking through on a Facebook ad, making a purchase.

Lytics analyzes user behavior using data science to summarize behaviors into practical scores that enable better targeting based on how customers interact with your brand. Read more on Behavioral Analysis.

Lytics provides six Behavioral Scores out-of-the-box to help you better understand and target your users. These scores are self-learning and self-maintaining, meaning they require no manual input and are automatically generated. Therefore, an audience created today using a score will be as effective in six months as it is today.

Blacklisting topics removes them from the Topic Taxonomy and allows additional topics to flow in. Certain topics generated from automated machine learning processes can be misclassified or lack domain knowledge or context, so it may be desirable to remove them.

Marketing tactics such as an "email campaign" or an "ad campaign".

Characteristics are a type of pre-defined audience made up of user fields or a combination of fields that enhance segmentation capabilities. Examples include "has email", "active in the last 90 days" etc. Read more in this FAQ.

The process of setting up an account and training the customer on how to use the core functionalities of Lytics to achieve their marketing goals and use cases.

Client-side Integrations are implemented in browser-facing code. These are implemented through the Lytics JavaScript tag and usually require you to have a tag from the third-party provider also installed on your web page.

Connected Customers are users who are engaged across multiple channels and are more likely to continue engaging with your brand. Lytics introduced the Connected Customer KPI as part of the Intelligence Reports.

An Integration that has been authorized between Lytics and a third-party tool.

Content Affinity is a measurement of a user’s interest in a particular content topic. It can be loosely interpreted as the probability that a user will interact with content about that topic.

Lytics Content Affinity Engine is a programmatic approach to extracting meaningful topics from URLs or other sources of content, and calculating users’ affinities based on their interactions with that content.

The task of assigning one or more topics to a document (a URL, an email, ad, or other content type) as part of the content enrichment process. Lytics can also ingest existing topics and taxonomy. Read more about the Content Classification dashboard.

A Content Collection is a group of content (documents). Collections can be dynamic and rule-based, or static and hand-selected groups of content.

The Lytics Content Collection Builder is used for building dynamic or locked content collections, which are essential for setting up content recommendations.

Lytics Content Recommendations use Content Collections to deliver relevant, personalized content to users. Lytics uses the calculated affinities to match content to users. Recommendations can be delivered in various ways including emails, modals, on webpages, etc.

A content topic is a high-level category that summarizes the content of a given document. Topics are different from keywords in that keywords are extracted directly from the content’s body, whereas topics are extracted using statistical topic models that can often provide more contextual information.

• For example, a product review for shoes might mention “laces” and “foot” without ever mentioning the word “shoe”. A topic model can infer that the review is about shoes.

The number of users who have entered the audience chosen as the conversion event for a stage.

A conversion event is the “target” audience that is assigned to a Stage within a Lytics Journey. For example, if you choose “Known Email” for the conversion event, the stage's source audience will be users who do not have a known email.

The conversion rate for a campaign is the proportion of people who engaged with the call-to- action, compared to the number of people who were reached. During a campaign, the conversion rate will stabilize and become more accurate as the size of both groups increases.

Crawling or "seeding" content involves starting with a particular URL and crawling its links. Note this is different from Scraping Content.

Custom Rules within the Lytics audience builder allows you to add any rule based on a user field to the audience being built or edited.

The process of bringing raw data into Lytics from data sources, which can then be associated with mapped user fields for creating audiences, understanding behaviors, or level of engagement.

Automated feature extraction, which Lytics does by turning raw user data into meaningful behavioral scores and content affinities.

The process of importing or ingesting data from third-party tools and sources into Lytics.

Aligning machine learning techniques to better understand business outcomes, solve problems, and drive objectives. The Data Science capabilities built into Lytics provide Behavioral Scores and Content Affinities out-of-the-box.

Data streams are silos of raw data containing key-value pairs organized by source that can be used in the mapping of user fields. All data sent to Lytics must be sent through a Data Stream.

Predictive models Lytics uses for Delivery and Experience Optimization.

Delivery Optimization is a feature that automatically determines when to deliver messages to individual users using the built-in third-party Experience providers on the Journey Canvas.

A process summarizing past data that can be used for further analysis. These models don’t predict anything on their own, but serve as building blocks for predictive models, such as Lytics Behavioral Scores.

A document represents webpages, items, or products that a user has interacted with. In Lytics, these documents most commonly represent articles, blog posts, ads, emails, products, and videos.

The relationship between things in a graph data model. As part of Identity Resolution, Lytics creates an “edge” between two “entity fragments” (nodes) when we observe identifiers for two different fragments in the same piece of data, like a page view that has both a cookie ID and an email address. For content, topics in our topic graph will have stronger edges as those two topics co-occur more frequently in content.

An Enrichment Integration uses a third-party service to enhance existing Lytics user profiles with additional data such as information on the user's company (for B2B marketing), demographics, and social data. Integrations such as Full Contact and Clearbit have enrichment workflows.

• User Entity: An entry in the user table. The entire collection of data Lytics has regarding a specific user.
• Content Entity: A document in the content table.

Within Lytics, Experiences are marketing tactics that can be executed in different channels depending on the provider. Examples include an "email campaign", a "Facebook AdSet", “SMS message” etc. You can create personalized web modal Experiences directly in Lytics, or import existing campaigns from supported email and advertising providers.

Export Integrations send user profile data from Lytics to a third-party tool, usually at the audience level.

The process of transforming raw data into usable inputs for machine learning models.

Frequency is a behavioral score Lytics calculates using data science to measure how consistently a user is interacting with your brand over time.

A type of Lytics Experience that can export audiences to any Lytics supported integration that is not natively supported by the Journey Canvas.

A high-level objective that measures marketing success. Lytics Journeys are tied to a goal, such as increasing conversion rate, decreasing customer attrition, and increasing customer LTV.

Generally, graphs represent relationships between things, where the “things” are called “nodes” and the relationships are called “edges.” Lytics uses graphs to represent the connection between entities: a graph of users, a graph of content, a graph of organizations, etc.

No glossary items.

A collection of user context, behavior and content affinity data associated to a single Lytics User ID.

The process of creating complete user profiles by combining various data points about an individual (identity fragments) based on unique identifiers. See Stitching User Profiles and Identity Resolution for more information.

Part of the data onboarding process that involves setting up a Lytics account and providing initial training for the client.

Import Integrations ingest data from third-party tools into Lytics. This data will differ based on the provider and workflow. It can be anything from identifying information, demographic data, subscription status, or activity data for a user.

Lytics Insights provide visibility into the performance of your audiences and campaigns and allow you to activate that data in real-time. Insights are the combination of a fact and action, powered by data science.

Integrations connect Lytics to third-party marketing technology tools. Integrations are primarily import and export workflows, but also include data enrichment workflows, webhooks, and more. See Integrated Marketing Tools for more.

Intensity is a behavioral score Lytics calculates using data science to measure the depth of a user's typical interaction with your brand.

The Lytics JavaScript Tag (JS Tag) provides anonymous tracking of user behavior during their site visits, which is a crucial input for constructing user profiles and enriching them with behavioral scores and content affinities. The JS Tag is also responsible for associating the current visitor with their full user profile in Lytics, known or anonymous.

Within Lytics, Journeys are a framework for moving your customers through their lifecycle and driving them towards your brand’s top-level marketing goals.

The Journey Canvas is the UI view of a Lytics Journey that allows users to manage Stages and Experiences all from one place.

No glossary items.

The difference in conversion probabilities. In Audience Discovery reports, lift represents the degree to which users in the source audience are more likely to enter the target audience.

Lytics Lookalike Models enable you to discover users who are similar to your best customers, and create Predictive Audiences based on your model scores that will drive conversions and increase engagement across your channels. Lookalike Models have replaced the feature formerly known as “SegmentML.”

The Lytics Query Language (LQL) is used to define the transformation of uploaded records and event data into user fields on a customer profile. It transforms row-level event data into document-oriented user info.

The initial, self-service steps of the Lytics View package necessary to access the Intelligence Reports. The 4 steps involve selecting a use case, installing the Lytics JavaScript Tag, importing users, and importing campaigns.

Machine Learning (ML) is utilizing data to train computers/machines/processes to perform complex tasks and aid in making complex decisions. Lytics' features such as Lookalike Models, Predictive Audiences, and Content Affinities are powered by ML.

Measurable values used to demonstrate the effectiveness of campaigns across marketing channels. Common metrics include:

• ROI (Return on Investment)
• CPA (Cost Per Action or Cost Per Acquisition)
• ROAS (Return On Advertising Spend)
• CLV (Customer Lifetime Value)
• CRR (Customer Retention Rate)

A set of data that describes and gives information about other data.

Metadata on a web page that contains structured information about common properties, like descriptions, titles, authors, thumbnails, etc. Meta Tags can help search engines and social media platforms know how to present content previews to users and are an important component of any content strategy.

The output of customer data and a machine learning algorithm that can provide predictions on the data.

Momentum is a behavioral score Lytics calculates using data science to measure the rate at which users are interacting with your brand. Momentum scores compare the user’s previous level of engagement to their present level of engagement.

Natural Language Processing (NLP) is a programmatic approach to processing and analyzing large amounts of natural language data. NLP is built into the Lytics Content Affinity Engine.

Lytics’ optimization can determine, from a set of available marketing tactics or experiences, which experience is most likely to result in a user converting on a goal or an action. This selected experience is often referred to as Next Best Experience and can change in real-time with user behavior.

Predicted time of a user’s next visit. Technically speaking, this is when we’d expect to see a user next on a particular data steam.

A “thing” in a graph, which for Lytics is typically a user or a piece of content.

Models that are updated in a batch such as email data.

Models that are updated continuously, in real-time. For example, identifying people onsite who are likely to abandon their cart.

Lytics Orchestrate is the legacy package name for what is now called the Lytics Journey Canvas. Orchestration focuses on automating and streamlining cross-channel customer journeys at scale.

Pathfora is a lightweight SDK (software development kit) built by Lytics for serving personalized modals on web pages via the Lytics JavaScript Tag.

Out-of-the-box audiences Lytics provides that package our data science scores into universally useful, marketer-friendly terms such as “highly engaged users”.

The output audience(s) that are built using the predictive score from a Lookalike Model.

Permissions determine what actions a user can take within the app, as defined by the user’s role. See Managing Account Users for details.

Personalize is a legacy Lytics product that allows you to serve web campaigns to users within different audiences. Lytics Experiences replace the functionality of Personalize Campaigns along with new features such as Delivery Optimization.

Propensity is a behavioral score Lytics calculates using data science to predict how likely a user is to return or engage with subsequent activity.

A third-party tool that Lytics connects to such as Facebook or Salesforce. Also known as an "integration".

Quantity is a behavioral score Lytics calculates using data science to measure a user's cumulative activity over their lifetime of brand engagement.

Data that has not been processed, modified, filtered, or transformed in any way.

The number of unique individuals who have seen a given Experience or Campaign.

Instantaneous, or near instantaneous, processing of data.

The reprocessing of all raw data stored within a Lytics customer’s account. This typically is needed after a change in identity resolution (for example, a new or changed unique identifier) or Lytics Query Language.

Recency is a behavioral score Lytics calculates using data science to measure how recently the user has visited or engaged. Recency scores compare the user’s previous level of engagement to their present level of engagement.

A list of “events” processed by Lytics behind the scenes including details on newly classified content and changes in audience segment membership.

User roles define the amount of access and permissions a Lytics user has when they are logged in the app and through the Lytics APIs. See Managing Account Users for details.

The schema defines how you map event data to Lytics. Each Lytics table has its own schema. For example, a simple user schema could be created for the fields email, name, and address.

The Schema Audit feature within Lytics UI provides an overview of how user fields are being used, the data they contain, and where that data is coming from (which data streams).

• Frequency: visitor has engaged more consistently than the average visitor.
• Intensity: visitor has engaged more deeply than the average visitor.
• Momentum: visitor has engaged at a greater rate than the visitor has in the past.
• Propensity: visitor is more likely to be engaged in the future than the average visitor.
• Quantity: visitor has engaged more times than the average user.
• Recency: visitor has engaged more recently than the average user has in the past.

The output of Lytics’ models comprised of under-the-hood calculations, typically referencing our Behavioral Scores, which are listed above.

The process of retrieving content from URLs that Lytics ingests from users’ activity.

The process of a machine learning model retraining itself when new data comes in. For example, when a new topic emerges on your website, the Lytics Content Affinity Engine will dynamically extract topics and add them to content collections from models that were previously trained.

A group of entities: users, content, and others (accounts, organizations may be added to Lytics). Segments are most commonly referring to audiences (a collection of user profiles to be targeted).

• In the context of Lytics Segment API, segments are named, logical expressions of users.

Segment Collections are grouped/named lists of segments.

The process of extracting every user of a segment (aka "audience"), which is used in most export workflows.

Server-side Integrations allow Lytics and the third-party service to communicate directly, usually in the form of a workflow. These workflows are available to set up and execute in the Lytics platform under Integrations.

A required input for building Lytics Lookalike Models, this is the group of users you want to reach with your marketing messages to get them to convert on a particular marketing goal. For example, “unknown users.” Lytics calculates a score from 0 to 1 representing the user's likelihood to convert to the target audience.

An Experience that is not in a Stage therefore not in a Journey. Standalone Experiences can be used for one-off marketing efforts when a full Campaign/ Journey is not necessary or desired.

Stages are the individual steps that make up a complete Journey. Users leave a stage when they complete the conversion event such as capture email, make a purchase, etc.

A subscription within the Lytics platform is how a work describes which events it expects to receive. It is defined as a query with SQL-like syntax, and is commonly used to consume segment events and system events.

A required input for building Lytics Lookalike Models, this is the group of users that represents the desired outcome for users in the source audience (have converted in the past). For example, “users with email addresses.”

Events responsible for triggering actions on workflows. These events are passed to a work if that work has a subscription for them. Such events can be audience events (where a profile joins or leaves an audience) or system events generated when a work changes state to completed, sleeping, running, etc.

The Lytics User ID or cookie ID is a primary unique identifier collected from the Lytics JavaScript Tag.

A Unique Identifier is a special type of user field that Lytics uses to match and merge user data across different data sources. The two most common examples are UID and email address.

The collective actions, strategies, or tactics executed to achieve a business goal. For example, delivering content recommendations in newsletters to drive increased engagement based on users' affinities.

An individual person using a product or service.

User Fields are a filtered and aggregated view of your raw user data.

A User Profile is a living record of an individual that interacts with your brand. In Lytics, all data gets processed as users fields, which then become attributes on a user profile.

• Technically speaking, a user profile is a unique combination of one or more identity fragments. If any one of the identity fragments in a given profile resolves to an identity fragment in another profile via a common unique identifier, then Lytics will merge these profiles together to create a rich single user profile for an individual.

UTM tracking codes are parameters that you add after the "?" in a URL. When a user clicks on a URL with UTM parameters, those parameters are sent back to your Google Analytics for tracking. Read more on using UTM Parameters for Attribution and Audience Building.

No glossary items.

A process that sends or receives data from Lytics to a third-party tool. Typically referring to an “integration workflow” such as exporting users from Lytics to Facebook.

Whitelisted Fields are an account setting option that will surface user fields to the browser via the Lytics JavaScript Tag to unlock various web personalization use cases.

No glossary items.

No glossary items.

No glossary items.