Marketing BS Detector
Scans a report, deck, or pitch for vague claims, vanity metrics, and unsupported projections - quoting each one and naming what real evidence would need to look like.
Get This Skill
Run this command in your terminal, or paste it into Claude—or your favorite AI assistant—and ask it to install the skill for you.
What This Skill Does
Names exactly what's unproven
Quotes the specific phrase making the claim, not a vague sense that "something feels off."
Separates volume from value
Flags a vanity metric (impressions, followers) that never actually ties back to leads, pipeline, or revenue.
Never overreaches
States a claim is unsupported, never that it's false - the difference between "unproven" and "wrong" stays intact.
Tells you what would actually back it up
Names the specific evidence - a number, a comparable case, a stated method - that's currently missing.
Best Used For
- Sniff-testing an agency update or pitch before taking its claims at face value
- Checking whether a "big win" in a report is actually tied to a business outcome
- Separating a real result from a confident-sounding one before a renewal or budget conversation
- Getting a second opinion on a forecast or projection with no stated basis
How It Works
Marketing material almost always sounds confident - a report, an agency update, a pitch deck - and confident language is easy to mistake for substantiated language. A vague claim, a vanity metric, or an unsupported projection can all read just as smoothly as something genuinely backed by evidence.
This skill reads the material you paste in and flags exactly where that gap shows up: it quotes the specific phrase, categorizes the pattern - a vague claim, a vanity metric, an unsupported projection, or hand-waving language - and states what real evidence would need to exist to back it. It never treats a definable term the same way as a claim - only claims get flagged here.
You walk away with a clear list of exactly what the material doesn't currently support - never a verdict that anything is false, just an honest read of what's proven and what isn't yet.
Get This Skill
Run this command in your terminal, or paste it into Claude—or your favorite AI assistant—and ask it to install the skill for you.
Want Marketing BS Detector Custom-Built for Your Team?
Marketri can adapt this skill to your brand, data, and workflows — or build a new one from scratch.
Common questions
No - it tells you the material doesn't currently support the claim. Unproven and false are different, and this skill only tests for the former.
No - a term with a meaning but no truth value (like 'programmatic') isn't a claim to flag. Marketing Jargon Decoder handles those.
You can, but Marketing Proposal Reviewer is the fuller tool for that - it checks scope, contract terms, and pricing alongside language, not just the language itself.
Any real marketing text - a report, deck, agency update, or pitch. It needs actual text to scan, not just a topic description.
Subscribe to our Newsletter
By subscribing, you agree to receive marketing emails from Marketri. See our Privacy Policy.