For the complete documentation index, see llms.txt. This page is also available as Markdown.

Core concepts

The ten ideas the platform is built on, in the order the loop uses them.

Ten ideas, in the order the loop uses them. Each is defined here once, and the product sections describe where you meet it.

Context

Everything the system knows about your business, held in the Growth Brain: your audience, brand, and competition, seeded from public information at onboarding, confirmed by you, enriched by your connected tools, and sharpened by every outcome. Every claim carries its source, and what you say always outranks what was inferred. See Context and memory.

Goals, KPIs, and targets

A goal is a standing ambition, one of six, and it never expires. A KPI is what the business can actually read: if no connected source makes a number readable, it is not a KPI for your workspace. A target is an achievable, computed commitment under a goal: a number and a date. Outcomes move targets, and targets move KPIs. See Goals, KPIs and targets.

Identity

One person is one record, one company is one account, no matter how many fragments arrive. Verified matches earn an immutable OneSource ID. Everything else attaches as an alias. Identity is what lets an anonymous click and a closed deal turn out to be the same story. See Identity.

Audience

The accounts and contacts under management: watched, scored, and acted on continuously. Audiences are organized as cohorts under two segments, B2B and D2C. See Building and sharing audiences.

Decisions

The unit of work. A decision is a moment where the system chose one option over others, spent something finite doing it, and said what it expected to happen. Anything without all three is just a step. See Decisions and traces.

Plays

The motions. A play is a named, repeatable piece of work that moves a KPI: picked from the library, made yours, and running against your audience with runs and activations. Running a play produces decisions. See What a play is.

Pulse

The attention layer. The system works in the background and raises its hand only when something needs a human: an approval, a problem, a moment worth marking. See The queue.

Measurement

The honest scoreboard. Every event, decision, and outcome joins on one trace ID, so results are graded independently of what any channel reports about itself. See Traces and receipts.

Outcomes and learning

An outcome is what happened, measured against what the decision expected. The difference between the two is a learning, recorded so the next decision is better. This is why the system compounds instead of going quiet. See The compounding loop.

The loop

The shape that holds it all: capture, resolve, score, decide, attribute, then around again. Context feeds decisions, decisions produce outcomes, outcomes sharpen context. See The loop in one page.

Last updated