Google Shopping feed optimisation guide
How titles, identifiers, product types, images, availability and landing-page consistency affect feed quality.
Read guide →Product data operations
Turn messy product data into feeds that perform.
Import from files, ecommerce platforms, SFTP or APIs. Map, transform, enrich, validate and distribute channel-ready product data from one operational layer.
Autumn Catalogue
248,310 products · updated 2m ago92
FEEDGRADE
Approved228k
Warnings17.8k
Errors2.5k
ProductBrandGTINStatus
Trail Runner ProNorthline✓ ValidReady
Studio Knit HoodieArc & Co.MissingFix
Core Tech BackpackVektor✓ ValidReady
Motion Flex TeeNorthline✓ ValidReady
Shopify connected24,180 products synced
AI enrichment ready1,248 weak titles detected
4 outputs healthyGoogle · Meta · TikTok · Affiliate
Built around the feed workflow
Not another dashboard looking for a problem.
Feednetic is structured around the work product-data teams actually do: connect sources, join data, normalise it, apply rules, improve weak attributes, diagnose quality and publish clean outputs.
See the platformOne catalogue in. Every channel ready.
Connect
Files, stores, SFTP and APIs.
Transform
Field mapping, joins and rule logic.
Improve
Quality checks and AI enrichment.
Distribute
Channel-ready outputs and schedules.
Four stages. One data layer.
Explore workflow
StoreShopify UK
FileSupplier XML
SFTPWarehouse stock
APIPricing service
FileBrand metadata
StoreWooCommerce B2B
product_name
→title
manufacturer_code
→mpn
stock_status
→availability
main_image
→image_link
IF brand
equals
Northline
THEN title
prepend
Northline |
IF product_type
contains
Footwear
THEN custom_label_0
set
High intent
Google246,801 ready
Meta247,210 ready
TikTok244,990 ready
Affiliate248,310 ready
SFTPNightly export
APILive endpoint
Go deeper
One platform. Different jobs.
Rule BuilderExplore rules
Change thousands of products without touching the source file.
Build conditions, transformations and labels with predictable, inspectable logic. Rules are ordered, testable and visible before publication.
Conditional logic across any mapped field
String, numeric and availability transformations
Custom labels for campaign segmentation
Preview before applying changes
IF brand = "Northline"
AND product_type CONTAINS "Footwear"
THEN title = CONCAT(brand, " | ", title)
SET custom_label_0 = "High intent"
SET google_product_category = "Apparel & Accessories > Shoes"
✓ 18,420 products matched
AND product_type CONTAINS "Footwear"
THEN title = CONCAT(brand, " | ", title)
SET custom_label_0 = "High intent"
SET google_product_category = "Apparel & Accessories > Shoes"
✓ 18,420 products matched
AI EnrichmentExplore AI enrichment
Use AI where it improves data, not where it creates uncertainty.
Target weak attributes, generate suggestions with source context and keep human review where commercial risk is higher.
Title and description improvement
Attribute suggestions from product context
Taxonomy and category assistance
Approval workflow before publish
AI review queue
1,248 suggestions
Weak titles624
Missing colour310
Descriptions under 80 chars209
Category suggestions105
Feed HealthExplore feed health
Know what is wrong before the channel tells you.
Grade the source, mapped feed and channel output separately so teams can see where quality deteriorates and what to fix first.
Required and recommended field checks
Identifier, image and availability diagnostics
Before-and-after feed grade
Channel-specific issue views
Feed grade
NEEDS WORK82/100
Missing GTIN2,491
Weak titles7,804
Missing product type1,118
DestinationsExplore destinations
Publish the same catalogue differently for every destination.
Each output can have its own field rules, schedule and validation profile without duplicating the source workflow.
Google Merchant Center feeds
Meta catalogue outputs
TikTok commerce feeds
SFTP, URL and API delivery
Output status
5 active destinations
Google ShoppingHealthy
Meta CatalogueHealthy
TikTok Shop AdsHealthy
Affiliate XMLHealthy
Client SFTPHealthy
From data problem to usable feed
Show the issue. Show the fix. Show the effect.
Feed optimisation is easier to understand when the change is visible. These examples show the kind of product-data problems Feednetic is designed to identify, transform and validate before a channel receives the feed.
Title optimisation
Turn vague titles into useful product identifiers.
Men's trainer
Northline Men's Trail Running Shoes - Black - Size 9
Feednetic approach: combine approved source attributes using rules, preserve landing-page accuracy, then validate character length and variant consistency.
Google Merchant Center requires a product title and allows up to 150 characters. Titles should accurately describe the product and match the landing page.
Identifier quality
Do not invent identifiers to fill a blank field.
GTIN: N/A
GTIN: 5012345678912
Feednetic approach: join trusted identifier data, validate format and flag products where no reliable GTIN is available instead of guessing.
Google says incorrect GTINs can cause disapproval. Only submit a GTIN when you are sure it is correct.
Taxonomy
Create product types that reflect how the business merchandises products.
product_type: Shoes
Sports & Fitness > Running > Trail Running Shoes
Feednetic approach: map inconsistent source categories into a controlled product type structure and keep channel taxonomy separate where required.
Availability
Standardise stock language before it reaches a channel.
availability: yes
availability: in_stock
Feednetic approach: normalise source values such as yes, available or 1 into the required destination value and monitor conflicts with the landing page.
Image quality
Separate a usable source image from a channel-ready image set.
1 low-resolution image
Primary + additional + lifestyle image set
Feednetic approach: validate image URLs, detect missing assets, join additional media sources and map primary, additional and lifestyle images separately.
Campaign segmentation
Turn commercial data into advertising labels.
custom_label_0: blank
custom_label_0: HighMargin
custom_label_1: Bestseller
custom_label_1: Bestseller
Feednetic approach: join margin or sales data, apply clear conditions and populate advertising labels without changing the storefront catalogue.
Designed for complex catalogues
See the whole product-data journey, not four disconnected tools.
Source health, transformation logic, enrichment and channel output belong in the same operational view. That makes problems easier to trace and changes easier to govern.
Open demo workspaceStorefrontProduct source
Feednetic CoreMap · rules · enrich
GoogleShopping output
MetaCatalogue output
WarehouseStock join
Where feed problems actually begin
Product data touches merchandising, operations and paid media.
A feed platform should make those hand-offs easier. The data usually starts in ecommerce systems, gets supplemented by operational sources and ends up in customer-facing channels.
Joined data
Use supplemental sources without flattening everything into one manual spreadsheet.
Traceable changes
Know which rule or enrichment changed an output field.
Channel logic
Apply destination-specific rules only where required.
Export control
Publish on schedules, URLs, SFTP or API endpoints.
Autumn Catalogue
248,310 products · updated 2m ago92
FEEDGRADE
Approved228k
Warnings17.8k
Errors2.5k
ProductBrandGTINStatus
Trail Runner ProNorthline✓ ValidReady
Studio Knit HoodieArc & Co.MissingFix
Core Tech BackpackVektor✓ ValidReady
Motion Flex TeeNorthline✓ ValidReady
Shopify connected24,180 products synced
AI enrichment ready1,248 weak titles detected
4 outputs healthyGoogle · Meta · TikTok · Affiliate
Connect the stack
Work with the systems already holding your product data.
Feednetic is designed around multiple source and destination patterns rather than forcing every team into one import method.
Product-feed resource hub
Practical guidance for teams fixing ecommerce data.
These pages are designed around real search intent and common operational problems, so the site can grow beyond a product brochure and build authority around product feeds, Merchant Center and catalogue quality.
How to improve product titles without keyword stuffing
A practical framework for building accurate, variant-aware titles from structured product attributes.
Read guide →GTIN errors: what to fix and what not to invent
Why identifier quality matters and how to handle missing or invalid manufacturer identifiers safely.
Read guide →Questions
Before another feed platform gets added to the stack.
The platform is designed for teams that need more than a one-off export or a single-store plugin.
Yes. The model supports a primary catalogue plus supplemental sources joined by a shared key such as SKU, product ID or another mapped identifier.
Yes. The delivery model includes scheduled URLs, SFTP and API-based patterns in addition to manual exports.
AI is positioned as an enrichment layer for targeted attributes such as weak titles, descriptions, categories or missing descriptive values. Deterministic rules remain available for changes that must be predictable.
Yes. Shared transformations can happen at catalogue level, while channel-specific logic can be applied only to the relevant output.
See the workflow properly