Core Concepts

The container for everything belonging to one brand: its audience data, campaigns, funnels, content, connected integrations, and agents. You always have exactly one Brand Space selected, and it decides what every other screen shows. The Brand Space screen itself has two tabs — Brand Assets (the logo, imagery, and file library) and Brand Management (name, category, and brand settings). Switch Brand Spaces from the selector in the header; agency accounts see a client switcher there instead.

The team layer. A workspace is where people, roles, and shared settings live, and a brand belongs to a workspace rather than to an individual. Roles are Owner, Admin, Editor, and Viewer: Owner covers everything including billing and workspace deletion, Admin can manage team members, budgets, and API keys but not billing, Editor can create and edit campaigns and content, and Viewer is read-only. Team members can additionally be assigned to specific brands, so one person can be an Editor on one brand and a Viewer on another.

A named, saved definition of a group of people, built in Audience. When you create one you pick a segment type — Behavioral, Demographic, Purchase History, Engagement, Lifecycle Stage, or Customer Value — then choose criteria within it and optionally narrow by date range and minimum activity. Generation runs as a background job rather than instantly; when it finishes the segment appears with a size, a status, and a growth rate. Segments are the unit other parts of the platform target: a campaign aims at one, and an integration sync pushes one.

A piece of marketing activity aimed at an audience, managed from Campaign Manager. A campaign is where targeting, budget, schedule, and creative come together, and where results come back. Campaigns can require approval before they go live, in which case they appear in an approval queue rather than launching directly.

A multi-stage path a customer moves along, made of ordered stages — awareness, consideration, decision, action, retention — each containing steps such as a landing page, an email, an ad, a webinar, a call, a piece of content, or an automation. Funnels have a type (lead generation, ecommerce, webinar, SaaS trial, or custom) and a status of draft, active, paused, completed, or archived. They can be built from scratch, started from a template, or imported. Each stage tracks its own entries, exits, conversions, and conversion rate.

A single piece of creative in Content Studio — a social post, story, ad, ad creative, article, email, SMS, image, video, caption, carousel, reel, infographic, whitepaper, case study, press release, or landing-page copy. Every asset carries a status: draft, in review, scheduled, published, or archived. Assets can be linked to funnel steps and campaigns, and they are what an agent produces when it generates something for you to review.

Software that does marketing work on your behalf. En Garde has two kinds. Workflow agents are ones you build yourself as a flow on a canvas in the Agent Suite, then trigger, schedule, or fire from an event. Walker Agents are built-in specialists — paid ads, SEO, content, audience intelligence — that watch your data and propose actions with a confidence score rather than executing on their own. Both kinds appear together on the My Agents screen.

One execution of an agent or workflow. A run is manual, scheduled, or trigger-based, and moves through pending, running, paused, completed, failed, or cancelled. It records its start and end time, how many steps completed, failed, or were skipped, a progress percentage, and any error message. Runs are visible in Workflow Activity as a chronological log and in Executions as a monitor for in-flight and finished work.

A recorded touchpoint on a customer journey — an ad seen, an email opened, a page visited — which Attribution uses to assign credit for a conversion. Because a single conversion usually has several touchpoints, the credit depends on which model you read it through: first touch, last touch, linear (spread evenly), time decay (later touchpoints weighted higher), or position-based (first and last weighted at 40 percent each). Attribution reports the same events under several models side by side, so a channel can look different depending on the question you are asking.