Google Shopping Feed Audit Checklist for UK Ecommerce Businesses

Google Shopping Feed Audit Checklist for UK Ecommerce Businesses

Google Shopping performance often depends on work that shoppers never see: product data, Merchant Center settings, tracking, diagnostics, and the way Shopping or Performance Max campaigns interpret the feed.

This checklist is designed for UK ecommerce teams, developers, platform partners and marketing managers who need a practical way to inspect a Shopping feed before more budget is pushed into Google Ads.

Downloadable worksheet: Google-shopping-feed-audit-worksheet.xlsx

Who this checklist is for

Use this resource if:

• 
  products are approved but Shopping performance is inconsistent
• 
  Merchant Center diagnostics are being handled reactively
• 
  product titles are too short, too long or too generic
• 
  variants are showing with weak or duplicated attributes
• 
  pricing, availability or delivery information is causing trust issues
• 
  Performance Max has spend but not enough feed-level clarity

If you want support beyond this checklist, Byte Digital's Google Shopping agency page explains how feed work connects to campaign structure, bidding and ecommerce growth. For broader paid search context, see the Google Ads agency page.

Merchant Center setup checks

Start with the account foundations before editing product rows.

• 
  Business information is complete and consistent with the website.
• 
  Website URL is verified and claimed by the correct Merchant Center account.
• 
  Tax, delivery and returns settings reflect what users see on the site.
• 
  Account-level warnings and policy issues are reviewed before item-level issues.
• 
  Automatic item updates are understood, not left to mask feed problems.
• 
  Feed rules, supplemental feeds and content API updates are documented.
• 
  Users, linked accounts and notification contacts are current.

If more than one system updates product data, record the source of truth. A platform feed, a supplemental sheet and manual Merchant Center rules can quietly conflict.

Product title and description formulas

A weak title makes Google guess. A useful title gives Google and the shopper enough commercial context without becoming spammy.

For many ecommerce products, use a title pattern like:

Brand + product type + key attribute + model or range + size or quantity + colour or material

Examples:

• 
  Weak: Bracelet
• 
  Improved: Annie Haak Silver Charm Bracelet - Mini Orchid Collection
• 
  Weak: DJI battery
• 
  Improved: DJI Intelligent Flight Battery for Mini 4 Pro Drone
• 
  Weak: Roofing sheet
• 
  Improved: Coroline Bitumen Roofing Sheet - Black - 2m x 950mm

Descriptions should support the title with useful detail:

• 
  product use case
• 
  compatibility
• 
  dimensions
• 
  material
• 
  pack size
• 
  delivery or handling detail where relevant
• 
  warranty or guarantee if it helps the buying decision

Do not stuff keywords into titles or descriptions. The goal is clear product matching, not a search-results word cloud.

GTIN, MPN and brand requirements

Identifiers matter most when products are manufactured goods with recognised global identifiers.

Check:

• 
  GTIN is present for products that have one.
• 
  GTIN is not reused across variants.
• 
  MPN is present where GTIN is unavailable but a manufacturer part number exists.
• 
  Brand is the real product brand, not the retailer name unless the item is own-brand.
• 
  Identifier exists is set correctly, not used to avoid data cleanup.
• 
  Marketplace or bundled products have a documented identifier approach.

For ecommerce accounts similar to the Drone Safe Store case study, compatibility and model identifiers can be especially important because users are often searching for parts, accessories and upgrades around exact product ranges.

Product variant handling

Variants should help Google show the right item, not create a set of near-duplicate products.

Review:

• 
  item group ID links true variants of the same product family
• 
  size, colour, material, pattern or style attributes are populated where relevant
• 
  each variant has its own correct image, price, availability and landing URL
• 
  variant titles include the attribute that distinguishes the item
• 
  unavailable variants are not left live with misleading availability
• 
  parent products and child variants are not both submitted in a confusing way

If a shopper would care about the difference, the feed should make the difference clear.

Image checks

Images affect approval, click quality and buyer confidence.

Check:

• 
  image URLs crawl correctly and do not require blocked sessions
• 
  product is clearly visible on a plain or suitable background
• 
  image matches the submitted variant
• 
  no promotional text, watermarks or overlays are used where they may breach policy
• 
  image dimensions meet Google requirements
• 
  lifestyle images are used only where suitable and supported by clear primary images
• 
  additional image links show useful angles, packaging, use cases or scale

Pricing and availability consistency

Merchant Center compares feed data with landing page data. Mismatches create disapprovals, warnings and wasted click journeys.

Check:

• 
  feed price matches landing page price
• 
  sale price and sale price effective date are used correctly
• 
  VAT treatment is consistent
• 
  availability reflects the website
• 
  preorder and backorder products are labelled correctly
• 
  landing pages use structured data that agrees with the visible content
• 
  feed update frequency matches how often stock and price change

For larger catalogues, sample high-volume products, newly added products, sale items and products with frequent stock changes.

Shipping and returns information

Delivery and returns data should remove uncertainty.

Check:

• 
  delivery costs and thresholds are clear
• 
  delivery speed is realistic for the product and region
• 
  oversized, heavy or restricted items have accurate delivery rules
• 
  free-delivery thresholds match the website
• 
  returns policy is accessible from product pages
• 
  Merchant Center returns settings do not contradict the website

This matters for retailers with operational constraints, such as roofing supplies, bulky products or specialist equipment. The Alltype Roofing case study is a useful example of ecommerce where location, stock, fulfilment and product specificity all matter.

Diagnostics and disapproval triage

Do not treat diagnostics as a single queue. Separate the issues by risk.

Use this order:

1. Account warnings and policy issues.
2. High-revenue or high-stock products disapproved.
3. Products approved but limited.
4. Recurring warnings affecting a large share of the catalogue.
5. Low-priority formatting improvements.

For each issue, record:

• 
  affected product IDs
• 
  issue type
• 
  likely source system
• 
  first seen date
• 
  owner
• 
  fix applied
• 
  re-review date
• 
  outcome

This makes feed quality a maintained process rather than a monthly panic.

Conversion tracking checks

Shopping optimisation depends on reliable conversion data.

Check:

• 
  purchases are tracked once
• 
  transaction value is passed correctly
• 
  refunds and cancellations are handled in reporting where relevant
• 
  consent mode diagnostics are reviewed
• 
  enhanced conversions are configured where appropriate
• 
  GA4 and Google Ads attribution differences are understood
• 
  phone or lead actions are separated from ecommerce purchases if both exist

If revenue tracking is inflated, bidding decisions will usually follow the wrong signal.

Performance Max and Shopping campaign checks

Feed work should be reviewed alongside campaign setup.

Check:

• 
  asset groups are aligned to meaningful product groups
• 
  listing groups are not too broad for decision making
• 
  brand and non-brand behaviour is understood
• 
  search term and category insights are reviewed
• 
  underperforming product groups are not hidden by blended ROAS
• 
  best sellers have enough budget and impression share
• 
  seasonal products are prepared before demand peaks
• 
  feed labels are used for margin, season, priority or stock status where useful

Byte Digital's work with online retailers, including the Annie Haak case study, shows why Shopping performance is rarely just a bidding problem. Product data, campaign structure and commercial priorities have to support each other.

Weekly and monthly maintenance schedule

Weekly:

• 
  review new disapprovals and high-impact warnings
• 
  check price and availability mismatch trends
• 
  sample top-spend products
• 
  review search/category insights
• 
  inspect best sellers and products with rising spend but weak return

Monthly:

• 
  audit title patterns by category
• 
  review identifiers and variant handling
• 
  compare Shopping performance by margin or product group
• 
  update custom labels for seasonality, stock or commercial priority
• 
  review tracking quality and revenue accuracy
• 
  document recurring diagnostics by source system

Quarterly:

• 
  re-check delivery and returns settings
• 
  review feed rules and supplemental feeds
• 
  compare product data quality against best-performing categories
• 
  plan pre-season feed improvements

Anonymised feed optimisation example

An ecommerce retailer sells branded accessories. A sample product originally used a generic title:

USB charger

The feed had no model compatibility in the title, duplicated images across variants and inconsistent brand fields. Shopping traffic arrived, but many clicks were for incompatible products.

Improved title:

BrandName USB-C Fast Charger for Model Range X - 30W - White

Other improvements:

• 
  brand corrected to the manufacturer
• 
  MPN added
• 
  compatibility added to description
• 
  variant image corrected
• 
  landing page structured data checked against visible price
• 
  custom label added for high-margin accessories

The expected result is not simply more clicks. The aim is fewer irrelevant clicks, clearer matching and better conversion quality.

FAQs

How often should a Google Shopping feed be audited?

High-volume feeds should be checked weekly for diagnostics and monthly for deeper product-data quality. Smaller catalogues can usually work on a monthly rhythm, with extra checks before sales, seasonal peaks or platform migrations.

Should product titles be written for Google or for people?

Both. A title should give Google enough structured detail to match the product and give shoppers enough information to decide whether the click is relevant.

Is Performance Max a replacement for feed optimisation?

No. Performance Max can use the feed heavily, but it cannot fix weak product data. Better titles, identifiers, images and product grouping usually make campaign decisions easier to interpret.

Related Byte Digital resources

For help with ecommerce PPC, see Byte Digital's Google Shopping agency service page. Relevant examples include Annie Haak, Drone Safe Store and Alltype Roofing. For wider PPC strategy, see the Google Ads agency page.