Як виглядає аудит готовності до Google Merchant Center від UPLIFY?
A real (anonymized) audit of a store's readiness for Google Merchant Center moderation: not_ready verdict 58/100, 2 critical blockers (shipping page, product with a third-party trademark), a spec with acceptance criteria, and an 11-step checklist before submission. No promises of passing.
Merchant Center readiness audit: an honest "not ready"
A real pre-flight audit of a real store — fully anonymized. This is a sample of an honest verdict: you can't submit right now, here are the two blockers, here is the spec with acceptance criteria, here are the 11 steps to submission.
What narrows the conclusions — read these first
- Merchant Center item level is unavailable (product statuses, item-issues, attributes) — API timeout, a known pattern for Prom accounts. Closed by a 30–60 min manual reconciliation plan (§ below) — the blind spot is named, not silently omitted.
- MC account level read successfully — there are no account-level problems.
- No GSC access — the indexing picture was not checked.
- The catalog was checked selectively (7 product cards + 2 categories out of 600+ SKUs in a single category alone) — a full review is split out as a separate task.
Verdict: fix first — then submit
The store is a genuine, living business with a strong trust foundation: a real pickup address, working phone numbers, an "About us" page with business identity, thousands of reviews, a working checkout (≈365–450k ₴/mo in revenue per GA4), valid HTTPS, and full Schema.org/Product markup. The Merchant Center account already exists and has no account-level problems.
1. Shipping page — a skeleton
BlockerA list of carriers with no costs, delivery times, processing time, and no "address delivery". The moderator compares the shipping page with the MC settings line by line — this is one of the most common rejection triggers.
2. Product with a third-party trademark
BlockerBedding «Луї вітон…» — the brand is in the title and the SKU. Google's counterfeit policy is the harshest: the account gets banned, not the product, and there is no easy recovery path. Even a single item is a direct risk.
3. High risks (3)
HighNo email in the contacts; the privacy policy is platform-only (Prom), not a proprietary one; the catalog has not been fully checked for third-party trademarks and licensed prints (toys, character costumes, textile prints).
How the score is calculated — explainable, not a "Google score"
Start from 100, deductions by trust-perimeter areas. The verdict is set by blockers, not the score: as long as even one critical "blocks submission" finding exists — the verdict is not_ready regardless of the score.
| Area (max deduction) | Deducted | Why |
|---|---|---|
| Policies: shipping / returns / privacy (−20) | −15 | shipping without figures (−10) · privacy platform-only (−4) · returns without a refund timeframe (−1) |
| Product cards (−10) | −7 | product with a third-party trademark (−5) · outcome promises in descriptions (−1) · unchecked remainder of the catalog (−1) |
| Contacts (−15) | −4 | no email (address and phone numbers — present) |
| MC configuration (−15) | −10 | sections are not read by the read tool — conservative deduction until reconciliation |
| Site↔feed consistency (−15) | −6 | MC item level unavailable, reconciliation not performed |
| Checkout + SSL (−15) | −0 | HTTPS valid, checkout works (GA4) |
| Identity / "About us" (−10) | −0 | identity complete |
| Total | −42 | 100 − 42 = 58 |
Spec with acceptance criteria — an example of depth
Every finding in the report is an executable task with "done correctly" criteria and a note on platform feasibility (what the Prom admin panel does out of the box and what it doesn't). Below are the two most important, abridged.
🔴 Rewrite the "Shipping & payment" page · blocks submission
- The first item — "Address delivery across all of Ukraine" (physically performed by the carrier's courier — that's correct; the platform's wording "courier delivery" is not enough for the moderator).
- For each method — the cost ("from 70–80 ₴") and the delivery time ("1–3 business days").
- Order processing time (cut-off) that does not contradict the business hours on other pages.
- The figures on the page = the figures in the platform checkout = the figures that will be in MC. A discrepancy = a moderator's nitpick.
- Acceptance criteria: 5 checkbox items, including parity of the RU/UA versions. Feasibility: config (the admin-panel page editor).
🔴 Remove the product with a third-party trademark · blocks submission
- Delete completely, don't hide — on templated platforms, hidden product cards remain in the structured data and the feed.
- An alternative only if the fabric has no branded print: a full rename + replacement of the SKU and photos; if the print imitates the brand's monogram — deletion only.
- After the fix — a pause of ≥72 hours (3–7 days is better) before submission: the platform's structured data and Google's caches need to update.
- Acceptance criteria: the URL returns 404/410; a search across the site and the catalog export finds no brand; the pause has been observed.
The report continues in the same format: email in the contacts (and a verbatim match with MC), a proprietary privacy policy instead of the platform one, a full catalog review for third-party trademarks and licensed prints (cartoon characters on textiles — a typical trap; legal resale of licensed board games — not a problem), health-outcome promises in product descriptions, the refund timeframe.
MC settings: mirror the site, no discrepancies
The "Business information" and "Shipping" sections aren't read by this run's read tool — so the report honestly frames them not as "detected problems" but as reconciliation instructions for a specialist with access: name/address/phone/email verbatim as on the site; in shipping — one carrier (address delivery), processing time/delivery time/cost = the figures on the shipping page; verified + claimed active. Plus a 30–60 min manual feed item-level check plan (disapprovals, feed language ↔ URL language version, spot price reconciliation, identifier coverage).
11 steps to submission — with effort estimates
| # | Task | Owner | Time | Blocks submission |
|---|---|---|---|---|
| 1 | Rewrite the shipping page (address delivery + cost + delivery times + cut-off) | administrator | 0.5 day | ✅ yes |
| 2 | Remove the product with a third-party trademark | administrator | 15 min | ✅ yes |
| 3 | Full catalog review for third-party trademarks/prints | administrator (QA — UPLIFY) | 0.5–1 day | ⚠️ risky without it |
| 4 | Email in the contacts (and then in MC) | owner | 0.5 day | ⚠️ risk |
| 5 | Proprietary privacy policy | administrator | 0.5 day | ⚠️ risk |
| 6 | Remove health-outcome promises from product descriptions | administrator | 1–2 h | no |
| 7 | Detail the refund (timeframe + who pays for return shipping) | administrator | 1 h | no |
| 8 | Reconcile Business info in MC + verified/claimed | MC specialist | 1 h | ✅ yes |
| 9 | Shipping in MC mirroring the site (after step 1) | MC specialist | 30 min | ✅ yes |
| 10 | Manual feed item-level check | MC specialist | 30–60 min | ✅ yes |
| 11 | Pause ≥72 h after the fixes — and only then submit | manager (oversight) | 3–7 days | ✅ yes |
Audit limitations — named, not hidden
The MC item level was not read (closed by a manual plan); the agent did not go through the final payment step (indirect proof — 1,421 purchases in GA4); the catalog was checked selectively; the Google Business Profile was not checked; there is no GSC access. Every blind spot has a closure plan — and none is passed off as verified.
How this audit was done
The report was prepared by Nestor AI — UPLIFY's AI operator — using a methodology that grew out of real Merchant Center unblocking cases; a senior specialist verified every finding. We don't promise passing moderation — the decision is always Google's. Our job is to make sure that at the moment of submission not a single known trigger remains, and that whatever couldn't be verified is honestly named and has a closure plan.