Signals and Scoring

Learn how Rembrandt detects buying signals and turns them into actionable intelligence for your sales team.

What are buying signals

Buying signals are indicators that a company is actively researching solutions, experiencing a relevant problem, or showing intent to purchase. Rembrandt scans multiple data sources, matches activity against your signal definitions, and surfaces the results as signal matches on your accounts and opportunities.

How signals work

Signals in Rembrandt are organized in three layers:

1. Signal Frameworks: You define what to look for (the catalog of buying signals)

2. Agents: Automated scanners that run your frameworks against accounts

3. Signal Matches: The detected results, which roll up into evidence on opportunities

Domains → Frameworks → Signals → Agents run frameworks → Signal Matches → Opportunities

Everything under Configuration in the sidebar is visible to Owners and Admins only.

Domains

Domains group related signal frameworks by business area. Think of a domain as a market category or industry segment.

Where to find them: Configuration > Domains

Each domain shows how many frameworks, agents, accounts, and opportunities are associated with it.

Signal frameworks

A signal framework is the catalog of buying signals your agents look for, organized by domain.

Where to find them: Configuration > Frameworks

Creating a framework

1. Go to Configuration > Frameworks

2. Click Create Framework

3. Choose a domain and give the framework a name and description

4. You'll be taken to the Signal Editor to define your signals

Signal properties

Each signal in a framework has:

Signal stages

Stages represent where a company sits in their buying journey:

UNAWARE
AWARENESS
CONSIDERATION
OPTIMIZATION

Signal tiers

DIRECT
INDIRECT

Data sources

JOBS
NEWS
WEB
REPORT

Import and export

You can import signals from a CSV file or export your framework's signals to CSV for sharing or backup.

Agents

Agents are the automated scanners that run your signal frameworks against companies to find matches.

Where to find them: Configuration > Agents

What an agent does

An agent takes a signal framework and scans accounts looking for matches across the enabled data sources. When it finds activity that matches a signal definition, it creates a signal match.

Configuring an agent

When setting up an agent, you:

1. Select a framework: Choose which signal framework the agent should use. The framework tab shows how many of its signals target Jobs, News, and Reports, so you can see up-front which sources this agent will actually use.

2. Pick the accounts: Choose which accounts the agent scans.

3. Configure schedule: Set how often the agent runs (or run manually)

Agent runs

Each time an agent runs, it produces results you can review:

• Signal matches found during the run

• Accounts scanned

• Jobs, news articles, and reports analyzed

You can view run history and drill into the results of any run from the agent detail page.

An agent run carries one status while it works:

SCHEDULED
PENDING
RUNNING
PROCESSING_SUMMARIES
PROCESSING_OUTPUT
POST_PROCESSING
NEEDS_FEEDBACK
PARTIALLY_COMPLETED
COMPLETED
FAILED
CANCELLED

Signal matches

Signal matches are the detected results: specific evidence that a company is showing a buying signal.

Where signal matches appear

Signal matches surface as evidence on the opportunity they contributed to:

• Opportunity detail > Evidence Blocks: The evidence the opportunity is built from, with the signals citing each block

• Opportunity detail > Evidence History: How that evidence changed over time, one entry per regeneration

Evidence history

The history shows each generation of evidence in chronological order, so you can see how an account's buying signals, and its conviction score, have evolved over time.

Conviction scores

When signal matches accumulate for an account and pain cluster, Rembrandt calculates a conviction score that represents the strength of evidence.

The minimum conviction score an opportunity needs to enter the review pipeline is set by an Owner or Admin under Settings > Organization > Sales > Pipeline. It defaults to 4, so a score of 4 or higher is normally what puts an account in front of a reviewer in Sales > Funnel.

Pain clusters

A pain cluster defines a problem you solve and how Rembrandt should detect opportunities for it. Each pain cluster carries the prompt that shapes how signal matches are interpreted and scored, the signal domains it draws from, and. On its Playbook tab. The buying committee to look for.

Where to find them: Configuration > Pain clusters

Pain clusters work together with signal frameworks. The framework defines what signals to look for, and the pain cluster defines how those signals map to a specific problem you sell against.

The Playbook tab

Each pain cluster carries a Playbook with three things outreach uses:

• Personas: the buying committee to look for on accounts in this cluster. Until a cluster sets its own, it inherits your organization default. See the Decision-making unit.

• Copy direction: what to lead with for this pain, what to avoid, and the line to land. One brief for every channel.

• Proof points: who you helped with this pain and what changed for them, used to back up the claim that you have solved it before.

Tone and message structure are organization-wide rather than per pain, and live in Sequences and Copy.

Pain clusters were previously called Solutions. The concept is the same; the name changed throughout the app.

Accounts and opportunities

Accounts

An account is a company that Rembrandt is tracking. Accounts accumulate signal matches over time as agents scan for buying signals.

Where to find them: Intelligence > Accounts

The account detail page has Overview, Opportunities, Contacts, and Feedback tabs.

Opportunities

An opportunity is a specific potential deal: a combination of an account and a pain cluster, with a conviction score based on the signal evidence.

Where to find them: open an account and use its Opportunities tab, or use Sales > Funnel for the pipeline view across your territory.

How they connect

Account + Pain cluster + Signal Matches → Opportunity with Conviction Score

Notifications

Signal activity is summarised for reviewers through qualification digests, not per-signal pings. See Sales Funnel and Campaigns for how those accounts reach review.

Putting it all together

Here's the typical workflow:

1. Set up your structure: Create pain clusters and domains, then build signal frameworks within them

2. Define your signals: Add buying signals to frameworks with stages, sources, tiers, and categories

3. Create agents: Attach a framework to an agent and pick the accounts it scans

4. Run agents: Let agents scan accounts, manually or on a schedule

5. Review results: Check the evidence on the resulting opportunities

6. Prioritize: Use conviction scores to focus on the strongest opportunities

7. Act: Review the account in Sales > Funnel, then run the campaign from Sales > Campaigns

Need help

• Check Sales Funnel and Campaigns for reviewing accounts and running outreach

• See the Getting Started Guide

• Contact support at support@rembrandtagents.com