← Back to CampaignBrief
Help & Reference

How to use CampaignBrief

CampaignBrief monitors public narrative movement, phrase convergence, rumor spread, money-message alignment, and public influence patterns across news, ads, social, podcasts, TV, and campaign finance. This page explains how to read what you see and what every term means.

The core concept: narrative status, not raw volume

CampaignBrief does not show you how many times a word appeared. It shows you whether a narrative is gaining momentum — combining source diversity, velocity, phrase repetition, outside spending, and influence actor amplification into a single status score. A hundred mentions from one Facebook account is noise. Three mentions from a PAC mailer, a local blog, and a radio segment is a signal.

The six status levels from low to high concern:

Noise
Isolated mentions, single source, no cross-platform pickup. Standard background chatter. No action required — stay aware.
Hype
Multiple social mentions but shallow sourcing. Often one viral post or one local personality. Monitor for whether it reaches press or paid amplification.
Watch
Cross-platform pickup beginning. Appears in multiple source types. Does not yet have press coverage or paid money behind it. Add to staff awareness and begin factual prep.
Moving
Active amplification across two or more source types. May have press mention, PAC language overlap, or influencer pickup. Prepare a public response option; brief candidate and senior staff.
Escalating
High velocity, confirmed press pickup, phrase convergence detected. Likely has paid amplification or organized distribution behind it. Communications team should be on alert. Response posture decision required now.
Crisis
Dominant narrative across multiple media types. Active press inquiries likely. High outside-spending alignment or coordinated phrase pattern detected. Immediate campaign decision required on whether to respond, rebut, or redirect.

Dashboard modules

Command Brief

Daily top-3 priorities distilled from all monitored signals. What the campaign staff needs to know before the morning call. Includes recommended posture for each item.

Narrative Radar

All active narratives ranked by impact score. Filterable by category, status, and time window. Each card shows what changed, why it matters, phrase convergence flag, and money-message alignment.

Phrase Convergence

Identical or near-identical phrases appearing across multiple independent sources. The signature feature of CampaignBrief. A phrase moving from a PAC mailer to a blog to a radio segment in 48 hours is a convergence event — not coincidence.

Rumor Watch

Unverified claims circulating in public sources. Each rumor carries a verification status, spread path, press-pickup flag, risk level, and suggested campaign posture. CampaignBrief monitors rumors — it does not amplify them.

Money + Message

FEC filings, Meta Ad Library disclosures, and independent expenditure reports mapped to narrative themes. When outside spending language matches a phrase detected in news or social, CampaignBrief flags the overlap.

Influence Map

Public actors — local officials, reporters, bloggers, PAC operators, advocacy group leaders — and their role in carrying or amplifying narratives. Categorized by role: originator, amplifier, press gateway, fundraising signal, etc.

Evidence Room

Every signal in CampaignBrief links to its underlying public-source evidence: the article, the ad, the social post, the FEC filing. Evidence Room is the audit trail — share items with counsel, comms, or opposition research staff.

Phrase convergence — what it is and what it is not

Important: Phrase convergence means CampaignBrief observed repeated use of similar language across multiple public sources. It does not constitute evidence of legal coordination, conspiracy, or any unlawful conduct. Do not characterize convergence as coordination without independent legal and factual review.

Phrase convergence is detected when the same 3–8 word phrase appears in two or more independent public sources — news articles, social posts, PAC ads, mailers, podcasts, or broadcast segments — within a defined time window.

Three things make convergence meaningful:

  • Source diversity — the phrase appears across different source types, not just multiple posts on one platform
  • Velocity — the phrase is new or accelerating, not a longstanding common phrase
  • Independence — the sources are not obviously quoting the same original article (CampaignBrief attempts to exclude syndicated content)

A convergence score of 0–100 reflects source count (log scale), source diversity, time window concentration, and source credibility. Scores above 60 are flagged as Watch or higher.

Recommended posture guide

Each signal and rumor carries a suggested posture. These are starting-point recommendations — final decisions require judgment from your campaign leadership and counsel.

Posture Meaning Typical action
Ignore Low-quality noise. Responding risks amplifying it. Log it. No public comment. Check again in 24 hours.
Monitor Worth watching but not yet actionable. Add to daily brief. Brief one or two staff. Prepare a fact sheet internally but do not release.
Prepare Moving fast enough that response materials should be ready. Draft a holding statement, fact sheet, or surrogate talking points. Do not release yet.
Respond Active press pickup or broad public amplification requires a response decision. Decide on proactive, reactive, or redirect strategy. Brief candidate. Activate comms plan.

Data sources

CampaignBrief uses public sources only — no voter files, no private data, no synthetic content. See the full Data Sources page →

News & RSS feeds
AP, Reuters, Politico, The Hill, Roll Call, NPR Politics, Washington Post Politics, Political Wire, PolitiFact, FactCheck.org, and local RSS feeds. Crawled daily. Scored by keyword match to race watchlist.
Google News RSS
Targeted query-based crawl of Google News for candidate names, opponent names, geographic area, and key issue phrases. Free, no API key required. Updates daily.
GDELT Project
Global Database of Events, Language, and Tone. Covers worldwide news media including TV captions. Used for broadcast pickup detection and narrative velocity scoring.
FEC / OpenFEC
Federal Election Commission filings — independent expenditures (Schedule E), 24-hour IE notices (F24), committee activity, and candidate finance data. Updated daily from the OpenFEC API.
Meta Ad Library
Public political ad disclosures from Facebook and Instagram. Searchable by advertiser, keyword, and geography. Ad text is scanned for phrase convergence matches.
Bluesky (AT Protocol)
Public posts from Bluesky searched by candidate name, opponent, and key phrases. Engagement-weighted — posts with 100+ likes or 50+ reposts are flagged as high-significance items.

Access & subscription

CampaignBrief is a paid-only intelligence service. There is no free tier. Access is token-based — when you subscribe, an access token is provisioned for your account and stored in your browser. You can manage billing, update payment, or cancel at any time through the billing portal.

Your access token identifies your account. Do not share it. If you need to add team members or transfer access, contact tom@phase3ai.com.

View pricing plans → | Sign in →

Compliance and neutrality

CampaignBrief is neutral by design. The same system monitors any party, any candidate, any issue. It is a monitoring tool, not a persuasion tool.
  • All monitoring uses public sources only — no private voter data, no commercial data brokers, no scraped private messages
  • CampaignBrief does not generate campaign communications, ads, attack scripts, or persuasion copy
  • CampaignBrief does not target voters or produce audience segmentation
  • Phrase convergence findings require human review before any campaign action is taken
  • Rumor Watch shows posture — it does not amplify unverified claims
  • Users are prohibited from using CampaignBrief output to defame, harass, suppress voters, or spread disinformation — see Terms of Service

Questions or feedback?

CampaignBrief is built for practitioners. If a signal is missing, a source seems wrong, or you need help interpreting something, contact us.

Contact us