How to Verify Political Claims Before They Reach a Campaign Memo

A political claim is ready for a campaign memo when a second person can open a public record and find the language the sentence depends on. Until that check is done, the sentence is a research note. The opposition-research page states the operating rule in plain terms: rumors, anonymous tips, and unsourced social posts can start a question, and they stay out of the file as claims until a public record answers the question. This how-to is that check. The wider field manual is The Complete Guide to Evidence-Backed Opposition Research for Campaigns at /blog/the-complete-guide-to-evidence-backed-opposition-research-for-campaigns.

The homepage says that on a published OppIntell briefing, each claim carries a source URL, a source type, and a citation validation status. Source types include an FEC filing, a Secretary of State portal, a court record, an official campaign site, or a government press release. The validation status records whether the cited document still exists at that URL and still contains the cited language. A briefing claim that fails validation stays out of the public post. A campaign memo can use the same gate.

Match the claim to one person

Read the name in the draft against the identity block before you read the exhibit. The opposition-research page says the legal name, the office, the jurisdiction, the cycle, and any namesake belong on the first page, anchored to a Secretary of State filing or an FEC committee name. A ballot name and a committee name are often different strings. If the memo uses a nickname or a headline name, stop and write the filing name next to it. Two candidates in one state can share a surname. A case caption, a donor name, or a vote list that matches only the surname is not yet a claim about this candidate.

The methodology page deduplicates the candidate universe on legal name, jurisdiction, office, and filing identifiers, with a reliability score on each upstream source. When two sources disagree, the verification note should say which source was preferred and why, using that score rather than a blended sentence. A memo that says "records show" without naming the record has skipped this step.

Open the record the sentence cites

A citation is a URL plus a source type, and then a reading of the document itself. Named types on the opposition-research page and the homepage are an FEC filing, a Secretary of State portal, a court record, an official campaign site, and a government press release. Open the URL. Confirm the document is the one named in the draft: the right committee, the right report period, the right bill, the right docket. Then confirm the cited language is actually in the document. The methodology page stores that result as citation_validation_status. A claim that fails does not flow into a public post. Pull the sentence out of the memo until the source is re-resolved or replaced.

Amendments change what "the record" means. The opposition-research page says amended finance reports replace earlier totals, so the file should cite the report that is current. A memo that quotes a receipt total from a superseded report is citing a document the committee has already replaced. Note the amendment date in the verification line. The public facts page describes a receipt for a published finding that includes supersession history. A memo line can carry the same note: which report was read, and which later report replaced it.

Quotes need a second look inside the same step. Store the words that were said, with the date and the place. A newspaper summary, a staff recollection, and a clipped caption are descriptions by someone else. The opposition-research page says the file should label them that way. A memo that presents a paraphrase as the candidate's sentence has failed verification even when the article URL still loads.

Keep the forms from collapsing into one sentence

Several public records sit near each other and answer different questions. FEC filings identify the committee, the treasurer, the reporting period, and itemized transactions above the disclosure threshold. They support sentences about who gave, when the report was filed, and whether a later amendment changed the totals. A state or local candidate may never appear in an FEC report. Disclosed money for that candidate lives in a state or local system. Personal financial disclosures list income, assets, and liabilities in ranges set by the filing rules. They are a different form. The opposition-research page warns that mixing them produces sentences that sound precise and cite the wrong document.

Donor concentration is a view of those finance filings, not a new record. Itemized contributions, PAC support, and recurring employers can be grouped by industry, family, or PAC. The map supports a sentence about the reports. A sentence about why the donor gave, or what the candidate promised in private, needs its own public source. A gift may sit next to a vote in the memo. Verification is done when both documents are cited and the sentence stops where the documents stop.

Votes have the same boundary. A vote line belongs when a legislative record exists for that person and that bill. A first-time candidate has no vote record to verify. The opposition-research page says a legislator's vote record exists only if they have served. Leaving the vote section blank is the verified result.

A practical pass on one sentence looks like this. Read the memo line and write the filing name it must match. Open the URL and note the source type. Copy the words or the figure the sentence depends on, and note the report or bill identifier. If an amendment exists, replace the figure with the current report and say so. If the sentence goes past those words, into motive or a private promise, move the extra clause into the question log. The line enters the memo only after that note is attached. The build that produces the file this check reads is /blog/how-to-build-a-citation-backed-candidate-research-dossier.

Use posture language while the record is still open

The methodology page separates two kinds of sentences. Where a public record supports a direct factual claim, the briefing states it and cites the record. Where a record opens a research question without resolving it, the briefing uses posture language: "may," "could," "would examine," and "appears consistent with." That wording is an editorial constraint the page publishes on purpose. A memo that upgrades an open thread into a flat assertion has not finished verification. Mark the sentence as a thread, or cut it, until a document closes it.

Contradiction flags, as the homepage uses the term, mark a statement and another public record that sit in tension. The flag is a prompt to read both documents before either one becomes the memo's sentence. It is not, by itself, a finding about motive. If the memo needs both, cite both. If only one document is in hand, the sentence stays a question.

The methodology page also says OppIntell does not human-review each public post before it appears. The trust model for the public site is the publish guard: content length, source-backed signals, structured data, and the absence of a broken citation or a source whose reliability score has dropped below threshold. Blog briefings regenerate hourly. Candidate research records refresh daily. When a cited source disappears, the post is held back until the source is re-resolved or replaced. A campaign memo is not that automated gate. The people named on the homepage's review step still read it: analyst, senior, legal, and communications. The verification standard they are checking is the same one the public pages describe.

Hold the line until review is finished

The homepage's spine is Capture, Govern, Analyze, Review, Export. Export names receipts, citations, and watermarking. A line waits at review until the URL, the source type, and the reading of the document are on the page. Writing the mail piece first and hunting for a citation afterward is the failure the opposition-research page describes: the desk discovers late that the citation does not say what the draft assumed.

The same check applies to the campaign's own candidate. The opposition-research page says a file that covers only the opponent misses the sentences an opponent can source from the same records. Verify those lines before they are needed in response, using the same identity match and the same document reading. The opposition-research page says this description is legal advice for no one. Recording a conversation, entering a system that is not public, and using personal data that a statute treats differently from a campaign filing stay with the campaign's counsel.

The use-cases page names other readers of the same records, including litigation, due diligence, communications, watchdog research, and partner vetting. Those labels describe work. They are not a client list. A campaign memo and a due-diligence memo can cite the same filing and still serve different decisions. The opposition-research page calls the non-campaign use public-record intelligence, at /public-record-intelligence. The source still has to be openable, and the sentence still stops where the document stops.

Public pages for the underlying records are /opposition-research, /about/methodology, /states, /races, and profiles under /candidates. The canonical coverage sentence dated 2026-10-03 is 24,000 U.S. political candidates in the public candidate-research index, all parties, federal, state, and local, deduplicated and index-gated. The homepage counter on 2026-10-05 read 24,130. Those are two published figures. This check does not merge them. For the full account of what the file covers, use the pillar at /blog/the-complete-guide-to-evidence-backed-opposition-research-for-campaigns.