OfferPrecheck is a paste-first check for promotional offer copy. You paste the wording before you publish, and it returns a scored pre-check: former-price language, discount arithmetic checked against the numbers you actually gave, and the phrases that tend to read as deceptive. I built it because "was $99, now $49" is the single most regulated sentence in marketing, and also the one founders write fastest.
The two traps it hunts
Deceptive-pricing enforcement mostly circles two things. First, the reference price: claiming a discount against a "former" price you never really charged, or against a price nobody paid. Second, the arithmetic: a "50% off" that is actually 30%, or a bundle "save $40" that saves less. Both are checkable with rules. The tool parses the numbers you provided, recomputes the claimed discount, and flags mismatches. It also scans for "was/now" framing and asks whether a reference price is even substantiated in the text.
This is arithmetic and string work, not judgment. A model will round "save up to 50%" into "half off" and call it fine. The recomputation is rules on purpose, because a wrong discount number is the one thing you want caught exactly. A model only drafts the plain-English summary.
What the report returns
Pre-check findings: the specific claim and why it fired. Fix checklist: tighten the wording, drop the unsubstantiated reference price, correct the math, or add the qualifier that makes the claim defensible. Decision-support framing: the score is a reason to rewrite, not a clearance to publish.
A low score means nothing obvious fired. It does not mean the promotion is lawful in every state or every channel — sale-price rules vary, and some jurisdictions add their own layers.
A concrete example: "Was $99, now $49, save 50%" — the arithmetic checks out, so the tool stays quiet on math. Change it to "save 60%" and it flags the mismatch immediately, because it recomputes from the two prices you gave, not from the claimed percentage. The reference-price scan is separate: if you wrote "was $99" but never sold at $99, the wording still reads as a former price and the tool asks you to substantiate it. Those are two different failure types, and keeping them separate is what makes the report useful instead of a single red blob. The OfferPrecheck page says this plainly, because the tool stops the careless draft, not the careful one.
What it cannot do
It reads the copy you paste. It cannot verify that a "former price" was a real, prevailing price in your market — that is a fact about your business, not your text. It does not know your state's specific sale-pricing statutes or the platform policies of the ad network you use. It is US-FTC oriented and text-only; it will not catch a misleading claim made in a product image.
So the output is decision-support. The report is the artifact you bring to a reviewer: "here are the two numbers to fix before this goes live". If a promotion is large or runs across regulated channels, the flagged items deserve a real compliance pass. The OfferPrecheck scan just keeps the obviously wrong math off the page.
FAQ
What does OfferPrecheck check?
Former-price wording and discount arithmetic — it recomputes the claimed saving against the numbers you provided and flags mismatches.
Does a clean score mean my promotion is legal?
No. It is a heuristic pre-check for decision-support. Sale-price rules vary by jurisdiction and channel.
Can it verify my "former price" was real?
No. It reads the copy. Whether a reference price was actually charged is a fact about your business.
Is this legal advice?
No. It is a heuristic pre-check. Confirm promotional-compliance questions with counsel.
