How to Review Marketing Claims

Marketing claims review is the process of reading campaign copy before it ships, extracting every factual or implied claim, checking whether each claim has matching substantiation, and rewriting risky language so the campaign keeps its intent without creating avoidable legal exposure. It is not a general copy edit. It is a claim-by-claim review of what the marketing says a product, customer, competitor, or outcome can do.

The work matters because risky claims rarely look risky to the team writing them. "Trusted by 10,000 teams", "the only platform for X", "secure alternative", and "faster than legacy tools" can all be reasonable or dangerous depending on the proof, the comparison, the jurisdiction, and the disclosure overlay.

Why marketing claims quietly create risk

Marketing teams move faster than substantiation files. Product capabilities change, launch copy drifts from the PRD, customer counts become stale, comparative claims pick up a named competitor, and a harmless-sounding endorsement may need a disclosure because the speaker has a material connection.

The dangerous part is usually specificity. Pure puffery is often low risk, but specific factual, comparative, implied, and absolute claims need evidence that matches the exact wording. A benchmark from last year may not support a current claim. A broad security posture may not support "HIPAA compliant" unless the contract, configuration, and BAA support it.

What the manual process looks like

Done by hand, a legal or product reviewer runs a disciplined sequence:

1. Collect the asset, campaign context, product notes, substantiation files, and launch deadline.
2. Extract every sentence or phrase that states a fact, makes a comparison, implies an outcome, or promises a specific result.
3. Classify each claim as vague or puffery, specific factual, comparative, implied, or absolute.
4. Test substantiation against the exact claim: who measured it, when, sample size, method, and whether it is apples-to-apples.
5. Research the currently operative advertising standard for the specific claim, such as FTC, NAD, state UDAP, sector rules, or platform policy.
6. Write a flagged memo with cut, rewrite, or substantiation-needed calls, then route it to an attorney before publication.

The time-consuming part is not reading the copy. It is keeping every claim, proof file, standard, and safer rewrite tied together so the attorney can approve the decision without rebuilding the review.

What an agent can automate

An agent can do the assembly and first-pass analysis while keeping legal judgment with the qualified reviewer:

• Extract and classify every claim. The agent reads the attached copy and separates pure puffery from specific factual, comparative, implied, and absolute claims.
• Check substantiation precisely. It tests whether the proof matches the exact wording, date, sample, and comparison basis instead of asking whether some proof exists somewhere.
• Research the operative standard. For each non-puffery claim, the agent looks up the standard that maps to that quoted claim and tags citations that need verification.
• Draft safer rewrites. The memo suggests language that keeps the marketing intent while reducing exposure, such as replacing unsupported absolutes with narrower, supportable claims.
• Verify before attorney review. A second verifier checks cited standards and substantiation calls so an untraceable citation does not reach the attorney as if it were settled.

The agent drafts the memo. It does not approve copy for publication, and it does not turn the memo into legal advice.

The guardrails that make it safe

The safe boundary is explicit: every memo waits for attorney approval before any copy ships. The agent can find claims, research standards, draft rewrites, and assemble the substantiation-needed table, but the publish decision remains a human legal act.

The second guardrail is verification. The bundle includes a claims verifier whose job is to reject fabricated, unmapped, or weakly supported standards before the memo reaches the attorney. That matters because a confident but wrong citation can be worse than no review at all.

Set it up in Task Machine

The Marketing & claims legal review playbook installs the legal review desk, the claim extraction and verification workflow, and the skills for marketing-claims review, launch review, feature-risk assessment, and legal-risk assessment. Setup takes a few minutes. You need a Task Machine workspace and permission to install playbooks (workspace owners have it). Web search and fetch access help the agent research public standards, but the workflow can still label research gaps when tools are unavailable.

1. Find the playbook

Open Playbooks in your workspace and search for "marketing claims", or browse the Legal category. The card shows the legal agent, verifier, review workflow, and skills the playbook creates.

2. Preview what it installs

Choose Preview & install to inspect the contents before anything is created. Review the Legal Agent, Claims Verifier, Claims Review Desk, Extract, check, research, verify, approve workflow, and the four assigned skills.

3. Define the review scope

Select Start setup and fill in the campaign or asset name, the claim types you expect, substantiation sources, and the review standard or jurisdiction. Strong inputs name the exact asset and proof sources, not just the campaign theme.

4. Generate and review

Use Generate customized playbook to bake the scope into the agent instructions and workflow prompts. In the review step, confirm that the workflow extracts claims, checks substantiation, researches standards, writes the memo, verifies citations, and waits for attorney approval.

5. Install

Click Install customized playbook to create the records. One follow-up lands in your inbox: start Extract, check, research, verify, approve. The first run asks for the marketing copy and substantiation files, drafts the flagged memo, verifies the cited standards, and waits for attorney approval before any copy is published.

What good looks like

A useful claims review makes the attorney's decision easier:

• Every claim is accounted for. The memo shows what was reviewed and why pure puffery was left off the flagged list.
• Substantiation matches the wording. The proof supports the actual claim, including date, sample, method, and comparison basis.
• Every cited standard is traceable. The citation maps to a quoted claim and survives verifier review.
• The rewrite is usable. The safer wording preserves the campaign intent while removing unsupported absolutes, comparisons, or implied outcomes.

Common questions

Does this replace an attorney review? No. The playbook drafts and verifies the memo so an attorney can review it faster. Publication approval remains with the attorney.

Can it review images or product UI claims? The bundle is built for words. It can flag an image that implies a claim, but visual review is a human call.

What if the standard cannot be confirmed? The agent should label the gap and stop rather than inventing a rule. The reviewer can broaden research, add source material, or escalate the claim.

Should every launch use this workflow? Use it when copy contains factual, comparative, implied, absolute, testimonial, influencer, atypical-results, auto-renewal, or regulated-sector claims. Pure brand lines with no factual claim may not need the full workflow.