Brand Voice Generator Catalog audit

Audit your whole product catalog in one run

Product descriptions are where risky claims live — prices, guarantees, “free from”, “100% natural”, “better than”, countdown urgency. Export your catalog to CSV from any store platform, upload it here, and get every row scored — no plugin, no integration, no developer.

Why product copy is worth auditing

Blog posts get read once; product descriptions carry your legal exposure at scale. The compliance ruleset’s most-triggered categories are exactly the phrases catalog copy is full of:

The same CSV can also be scored against your brand voice guide — same upload, different question: not “can we say this?” but “does this sound like us?”.

The five-step audit

  1. Export your catalog to CSV
    PlatformWhereColumns to keep
    Shopify Admin → Products → Export → CSV Title, Body (HTML)
    WooCommerce Products → All Products → Export Name, Short description, Description
    Magento / Adobe Commerce System → Data Transfer → Export → Products name, short_description, description
    Wix Stores Dashboard → Products → Export name, description
    Anything else Any product export that produces a CSV The name + description columns

    You don’t need to delete the other columns — the next step handles that.

  2. Choose the check

    Claims & compliance reviews legal and platform risk — no setup needed, works on any catalog. Evaluate scores each row against your brand voice; it needs a brand voice guide for your domain first. Many teams run both on the same file.

  3. Upload and pick the copy columns

    The column picker reads your file and pre-ticks the columns that contain prose, showing real values from each — IDs, URLs, prices and dates stay unticked. Tick up to 8 columns; title + description is the usual pick. Your selection is remembered for that export shape, so a weekly re-run needs no re-ticking.

    Descriptions exported as HTML are read as text. Light markup is fine, but each row is scored on its first 500 characters — if your descriptions are mostly tags, strip them first (a spreadsheet find-and-replace on <*> works).

  4. Set the row count and run

    Pricing is per 25 rows, and a batch that fails is refunded automatically — you only pay for rows that were actually scored.

    RowsCreditsNotes
    251The smallest paid unit
    1004Largest single compliance run
    50020Largest single brand voice run — one stitched report

    A brand voice evaluation covers up to 500 rows per run; a compliance review covers up to 100 rows per run — for a bigger catalog, run it in passes (sort the export so each pass is a stable slice, e.g. by product ID).

  5. Read, fix, re-run

    Every row gets its own score and flags, and the report says explicitly if any rows are missing (they’re the refunded ones). From there:

    • Refresh content — brand voice reports offer on-voice rewrites of the rows that scored 75 or below, worst first, shown beside the original so you copy what you want.
    • Per-flag fixes — compliance flags each come with a defensible rewrite of the risky phrase.
    • Download CSV — the row scores export back out for whoever owns the catalog.

Running audits for clients?

Team workspaces let an agency run this across a book of accounts: seats share one credit balance, work is grouped per client, seats can be restricted to their assigned clients, and each client can carry a soft credit budget — so a 500-row catalog audit for one client can’t quietly consume another’s retainer. See team workspaces in the help guide.

Approved a claim once (“money-back guarantee” with a signed-off basis)? Add it to the approved-claims register and every future catalog run annotates it instead of re-flagging it — with a reference and an expiry, per market.
Run a compliance audit Score against your brand voice Pricing

Platform export paths are current as of mid-2026 and may move between platform versions. The Claims & Compliance Checker is an automated drafting aid, not legal advice; flagged rows are candidates for human review, not verdicts.