How Sillage works
The key concepts behind Sillage in one video — from your persona and top accounts to the signals you act on.
Everything in Sillage builds toward one thing: a signal, a reason to reach out to an account right now. This video walks through every concept on one map, from what you set up to what you get back.
The three layers
Every concept in Sillage belongs to one of three layers:
Setup
What you configure: your persona, your top accounts, your agents and watchlists.
Content
What Sillage collects on your accounts: posts, comments, job postings.
Signal
What you get: a match between an agent you set up and the content we collected.
Setup: tell Sillage who you sell to
Your workspace starts with two things, set up through the app, the API, or the MCP server:
- Persona — the people you want to reach, described by job titles (RevOps, SalesOps, Head of Sales, ...) and location (EMEA, USA, anywhere).
- Top account list — the companies you want to win in the coming months, for example Salesforce, HubSpot, Slack.
As soon as both are in place, Sillage combines them to build your account mapping: for each top account, it finds the decision makers who match your persona. Add Salesforce with a RevOps persona in EMEA, and Sillage maps every RevOps at Salesforce in EMEA. See Account mapping for how matching works and what to do when an account cannot be matched.
Content: what Sillage collects
Once your accounts are mapped, Sillage starts collecting their activity. This is what we call content:
- Company posts published by each top account.
- Company job postings opened by each top account.
- Posts and comments on LinkedIn written by each mapped decision maker.
Content is the raw material. On its own it is just activity; the next layer is what turns it into something you can act on.
Agents: what to look for in the content
An agent watches your content for one specific kind of event. Sillage ships with around ten types — keyword detection, job updates, interactions with competitors, customers, partners, influencers or champions, keyword detection on job postings, and more.
Some agents need a watchlist: a list of entities to watch for, such as your competitors, customers, partners, influencers, or champions. Watchlists are part of your setup, just like your persona, and you can create your own custom ones.
Agents read and parse your content. When an agent finds what it was set up to look for, it creates a signal.
Signals: what you get
A signal is a match between the setup of an agent and some content. It is the end product of everything above, and the data you act on: read it through the API, sync it to your CRM, or feed it to an AI agent like Claude in your GTM motion.
Three examples of how agents produce signals:
- Keyword agent. You give it keywords to monitor, like "sales efficiency" or "buying intent". It reads all your content looking for them. Signal: Salesforce's Head of Sales wrote a post on sales efficiency on July 1st.
- Job update agent. It does not need posts or job postings — it only reads your account mapping and spots movement. Signal: John, Henry, and Francis joined Salesforce in the last two months — three separate signals.
- Competitor agent. You give it a watchlist of competitors, say Amplitude, PostHog, and BetterStack. It looks for interactions between those companies and your top accounts. Signal: Amplitude's Account Executive and BetterStack's CEO commented on a post by Salesforce's Head of Data.
Every time an agent finds a signal, it is added to your workspace, ready to use.