Amplitude Certification key terms

A single recorded action a user performs in a product, such as a button tap or purchase, sent to Amplitude with a name and optional properties describing when and how it happened.

A key-value attribute attached to a specific event that describes details of that action, such as plan type or item count, and can be used to filter, group, or break down charts.

A key-value attribute stored on the user profile rather than on a single event, describing persistent traits like country or signup date that apply across all of a user's activity.

The unique identifier assigned to a known user, used to stitch together events from the same person across devices and sessions so their behavior appears as one continuous journey.

An identifier generated for a browser or device that tracks activity before a user logs in, later merged with the User ID during identity resolution so anonymous and known activity combine.

An API call that sets or updates user properties and links a Device ID to a User ID, telling Amplitude who the person is and what traits should be stored on their profile.

A defined group of users who share a characteristic or behavior, such as signing up in the same week or completing a key action, used to compare retention and conversion across segments.

A chart that measures how many users complete an ordered series of events and where they drop off, revealing conversion rates between each step of a defined user flow.

A chart showing how many users return to perform an event after an initial action, typically displayed as a cohort table of returning users across successive time periods.

An analysis that counts users or events over time for a chosen event, optionally filtered and broken down by properties or cohorts to compare behavior across groups.

A chart setting that splits results into separate series based on the values of a chosen property, letting analysts compare metrics across categories like platform or country.

A measure of how frequently users perform a specific event within a time window, expressed as a distribution showing how many days or weeks users were active.

The process of receiving event and user data into Amplitude through SDKs, APIs, or integrations, including validation, enrichment, and eventual storage for analysis.

A credential that authorizes an application or server to send data to a specific Amplitude project, identifying which project should receive the incoming events.

A confidential credential used for server-side authentication with Amplitude APIs, granting broader access than the API key and therefore kept out of client-side code.

A container within an Amplitude organization that holds its own data, settings, users, and API credentials, typically corresponding to one product or application.

The documented, governed catalog of approved events and properties for a project, defining naming conventions and expected types so that analytics data stays consistent and trustworthy.

A specification describing which events and properties should be instrumented, their triggers and allowed values, used to align engineering and analytics before data is collected.

The policies and controls that keep analytics data accurate, consistent, and compliant, including naming standards, access restrictions, validation rules, and review of incoming data.

The total count of events ingested into a project over a billing period, which drives plan limits and costs and is reduced by filtering or blocking unwanted events.

An event type configured to be dropped during ingestion so it is not stored or billed, commonly used to suppress noisy, deprecated, or non-essential events.

A remote toggle that enables or disables a product capability for specific users at runtime, allowing teams to release code safely without redeploying the application.

A controlled test that assigns users to different variations of a feature and compares outcomes against a chosen metric to determine which variation performs better.

One version of a feature within an experiment or flag, such as control versus treatment, that a user may be assigned to and whose behavior is measured separately.

Rules that determine which users receive a feature flag or experiment variation, based on attributes, cohorts, or percentage rollouts rather than all users at once.

A measure of how likely an observed difference between variations is real rather than due to chance, typically expressed as a confidence level or p-value in experiment results.

The number of users or events included in an experiment, which affects the ability to detect a real effect and reach trustworthy conclusions about variations.

A recording of a user's interactions on a web or mobile interface, replayable as a video-like sequence that shows taps, scrolling, and navigation for qualitative debugging.

A privacy setting that hides or blurs sensitive content such as text inputs or personal data in session replays, protecting user information while preserving interaction context.

A user's permission to collect and record their behavior, which must be obtained where required and can be used to gate session replay or tracking until granted.

A visual overlay on a page or screen showing where users click, tap, or scroll most, using color intensity to reveal areas of high and low engagement.

A type of heatmap that aggregates clicks or taps by location on an interface, highlighting which elements attract attention and which are ignored or misclicked.

A visualization showing how far down a page users scroll and where they stop, indicating which content is actually seen versus dropped off before reaching.

An in-product message such as a tooltip, modal, or banner shown to users based on targeting rules to onboard, announce, or nudge them toward an action.

An in-product questionnaire delivered to targeted users to collect structured feedback, such as ratings or multiple-choice answers, tied to their behavior and profile.

The condition that determines when a guide or survey appears, such as a page load, a specific event, or a delay, controlling the timing of the in-product message.

Net Promoter Score, a survey metric derived from asking users how likely they are to recommend a product, calculated from promoter and detractor responses on a numeric scale.