Lock a brand's ad messaging hierarchy so every campaign stops reinventing the value prop from scratch

You are building the ad messaging hierarchy for this brand — the reference document a copywriter should pull from for every future paid campaign instead of re-deriving the value proposition from scratch each time.

WHAT THE PRODUCT ACTUALLY DOES
A payroll platform that runs multi-state tax filings automatically and flags compliance errors before they're submitted.

WHO IT'S FOR AND THEIR MAIN OBJECTION
Small business owners with under 20 employees; main objection is fear of switching payroll providers mid-year and breaking something.

WHAT'S BEEN TESTED SO FAR
"Save 10 hours a month" tested for 6 weeks, click-through was fine but conversion to trial was low.

COMPETITOR CLAIM TO NOT COLLIDE WITH
Gusto's ads lead heavily with "payroll made easy" — this brand needs to sound distinct from that.

Build the hierarchy in three tiers, most important first: the single primary claim every ad should be able to trace back to, three to five supporting proof points that back that claim with something concrete (a number, a mechanism, a named outcome — not an adjective), and a short list of phrases this brand should never use in ad copy, with the specific reason each one is banned. The primary claim must directly answer audience_and_main_objection — if the stated objection is about price, a primary claim about ease-of-use is answering a question nobody in this audience is actually asking, no matter how true it is. If tested_messaging_history shows a claim that underperformed, do not silently drop it — name it, name your best guess at why it likely underperformed, and say whether it should be retired or just needs a different proof point attached. The primary claim and at least one supporting point must be phrased distinctly enough that competitor_claim_to_avoid could not be mistaken for this brand's message if a prospect saw both back to back — if your first draft of the primary claim reads too close to the competitor's claim, revise it and note that you did.

OUTPUT FORMAT
1. Primary claim (one sentence).
2. Supporting proof points (3-5, each with the concrete backing evidence, not just the claim).
3. Banned phrases, each with the specific reason.
4. Verdict on any previously-tested underperforming claim: retire, or keep with a new proof point.
5. One line confirming the primary claim is meaningfully distinct from the competitor claim to avoid, or noting what you changed to make it so.