Case Study: How a Pet Supplies Brand Automated Feeds to Google Shopping, Meta, and eMag

A pet supplies company with 3,400 active SKUs was managing product feeds to Google Shopping, Meta Catalog, and eMag using a weekly manual export process. On a Tuesday morning in November, their Google Search Console showed a 64% overnight drop in Shopping impressions. The reason: a bad feed export had caused 200 product listings to be suspended for policy violations. The company had lost their primary paid traffic channel for the run-up to the holiday season.

The Challenge

The company’s product feed automation problem had been building for 18 months before the Google suspension made it impossible to ignore. They had started with Google Shopping only, using a simple CSV export from their WooCommerce store. When they added Meta Catalog and eMag, they kept the same process — one person exporting three separate files each Monday, formatting each manually, and uploading them to each channel.

Each channel had distinct format requirements:

• Google Shopping: Requires a specific set of mandatory attributes (id, title, description, link, image_link, price, availability, condition, gtin or mpn), with strict rules on price formatting, availability values, and title length. Product categories must map to Google’s product taxonomy.
• Meta Catalog: Similar mandatory field structure to Google but different title and description character limits, different category taxonomy, and a distinct requirement for product_type as a free-form categorization field separate from official categories.
• eMag: XML format rather than CSV, with a Romanian-language category taxonomy, mandatory attribute fields that vary by category, and strict requirements on EAN code presence for most product types.

The manual process required the same product data to be reformatted three times per week, with no systematic validation before upload. Errors reached the channels because the person doing the export had no mechanism to check the output against each channel’s rules before submission.

The specific failure that caused the Google suspension: a price update in WooCommerce for 200 products had included a currency symbol in the price field rather than a plain number (a WooCommerce plugin formatting change had introduced this). The manual export process did not strip the symbol. Google’s feed parser rejected the price format as a policy violation and suspended those listings. The suspended listings were not restored until the following Thursday — five days of lost Shopping exposure during the second week of November.

[SCREENSHOT: Google Merchant Center feed diagnostics showing 200 item disapprovals with the “Invalid price format” error message, representing the holiday-season suspension that triggered the automation project]

What They Tried First

The company’s initial response to the Google suspension was to add a manual verification step to the export process. The person responsible for feeds was asked to open the exported file in a spreadsheet and scan for obvious formatting issues before uploading. This added approximately 40 minutes per feed per week (120 minutes total across three channels) and caught two additional formatting errors in the following six weeks.

The problem with this approach was that it depended on the same person manually checking the same formats repeatedly, with no guaranteed coverage. A 3,400-row spreadsheet is not something a person can scan reliably for every possible format violation. The check caught visible formatting anomalies in obvious columns; it would not catch a systematic issue like incorrect availability values, missing GTINs on products that required them, or products that had been discontinued in WooCommerce but not removed from the active feed.

They also tried a third-party feed management tool — a well-known SaaS product specifically designed for Google Shopping feed management. It worked for Google Shopping but did not support eMag, which meant they still needed a separate manual process for that channel. Running two tools with different update schedules and two different interfaces created its own synchronization problems. When a product price changed, it had to be verified as updated in both systems before the weekly export.

The underlying issue remained: product feed automation was partial, manual, and fragile.

The Solution

The company moved their product catalog into MicroPIM and rebuilt their feed workflow around automated, scheduled exports with format-specific channel templates and pre-export validation.

Step 1: Feed Template Configuration per Channel

For each channel, the team built a feed template in MicroPIM defining:

• Output format: CSV for Google Shopping and Meta, XML for eMag.
• Field mapping: Which product attributes map to which required and optional feed fields.
• Value transformation rules: Price formatting (numeric, no currency symbol, two decimal places with period separator), availability values (WooCommerce “instock/outofstock” mapped to Google’s “in stock/out of stock” and eMag’s “available/unavailable”), title generation (for Google: brand + product name + key specification, max 150 characters; for eMag: product name + brand, max 100 characters).
• Category taxonomy mapping: A lookup table mapping internal WooCommerce product categories to the correct Google product taxonomy IDs and the corresponding eMag category paths.

The template build took 16 hours across four sessions — primarily the category taxonomy mapping work, which required matching 340 internal product categories to their correct equivalents in three different taxonomy systems.

Step 2: Pre-Export Validation Rules

Each channel template included a validation layer that ran before the export file was generated. The validation rules checked:

• Mandatory field presence: any product missing a required field (GTIN/EAN for products where it was required, availability value, price in the correct numeric format) was flagged.
• Value format rules: price as a numeric value without currency symbols, availability as one of the accepted string values, title length within the per-channel limit.
• Suspension risk detection: products that had triggered a channel rejection in a previous feed run were flagged automatically in subsequent validation passes until the underlying data issue was resolved.

The first validation run after migration flagged 89 products with issues: 43 missing EAN codes required for eMag compliance, 31 with price fields containing currency formatting artifacts, and 15 with title lengths exceeding Google’s limit. The team resolved all 89 in 3 hours before the first automated export ran.

[SCREENSHOT: MicroPIM pre-export validation report for the Google Shopping feed template, showing flagged products by rule type — missing GTIN (31), price format error (31), title length exceeded (15) — with a button to view each flagged product]

Step 3: Scheduled Automated Exports

With templates and validation in place, the team configured scheduled export jobs:

• Google Shopping: Daily export at 03:00, with automatic upload to Google Merchant Center via the Merchant Center Content API. Any products failing validation are excluded from the export and added to a notification email summary for the following morning.
• Meta Catalog: Daily export at 03:30, delivered to the Meta Commerce Manager feed URL.
• eMag: Daily export at 04:00, uploaded to eMag’s product feed endpoint.

The shift from weekly manual exports to daily automated exports had an immediate operational impact: price changes and inventory updates propagate to all three channels within 24 hours rather than up to 7 days. During the holiday season, this meant promotional price changes went live on Google Shopping and Meta the morning after they were set in WooCommerce, rather than waiting for the next Monday’s export cycle.

Step 4: Error Alerting and Recovery

The automated pipeline sends a daily digest report. When a scheduled export completes with zero validation errors, the report shows only a summary (products exported per channel, timestamp). When validation errors are present, the report lists each flagged product, the rule that failed, and the current field value. The catalog manager reviews this digest each morning, which takes 3 to 5 minutes on typical days.

For the notification triggered by an actual channel rejection (when a channel’s API returns errors after feed upload), MicroPIM sends an immediate alert rather than waiting for the daily digest.

[SCREENSHOT: MicroPIM automated feed schedule screen showing three configured export jobs with their schedules, last run times, product counts, and status indicators — all showing “Completed, 0 errors”]

The Results

The automated product feed pipeline has been running for seven months. In that period:

• Manual export work: Reduced from 3 to 4 hours per week (3 channels, weekly) to approximately 5 to 10 minutes per day reviewing the digest. Total weekly effort: under 1 hour.
• Feed-related suspensions: Zero since the automated pipeline was deployed. In the previous 12 months, there had been three Google Shopping suspensions and two eMag listing removals for feed errors.
• Data freshness on channels: Products now reflect current inventory and pricing within 24 hours. Previously, a product going out of stock on Monday afternoon would remain in the Google Shopping feed until the following Monday’s export, generating clicks on products that couldn’t be purchased.
• Holiday season Google Shopping performance: With no suspension period during November/December, Shopping impressions were up 34% year-over-year during the same period. The prior year’s 5-day suspension had occurred during the highest-spend window of the year.
• Feed compliance rate: 99.1% of products passing validation before export at last audit, up from the pre-automation state where the compliance rate was not measured but channel rejections were occurring regularly.

The feed automation project prevented a recurrence of the holiday suspension that had directly cost the company an estimated €8,200 in paid Shopping revenue (calculated from the average daily Shopping revenue in the days before and after the 5-day gap). The annual labor reduction — roughly 150 hours of manual export work per year — was a secondary benefit.

Key Takeaways

• Manual product feed automation is not automation: it is a recurring manual task with no systematic error prevention. The only way to prevent feed errors from reaching channels is validation before export, not review after upload.
• Weekly export cycles are too slow for modern ecommerce. Price and inventory changes need to propagate to advertising channels within 24 hours, not within 7 days.
• Scheduled automated feeds require per-channel format templates. A single product record needs to be transformed into three different formats simultaneously; trying to maintain this with manual exports or general spreadsheet tools does not scale.
• Feed validation errors that cause channel suspensions compound: a suspension during a high-traffic period (holiday season, promotional window) has disproportionate impact compared to the same suspension in a slow month.
• Daily digest reporting keeps humans in the loop without requiring manual work on clean days. The oversight burden is 5 minutes when everything is working.

If your team is still manually exporting product feeds to advertising channels, you are one formatting error away from a channel suspension. Product feed automation with validation is not a luxury for large catalogs — it is basic operating hygiene for any business running advertising on Google, Meta, or marketplaces. Set up your first automated feed in MicroPIM at app.micropim.net/register.

Related Reading

• Automate Everything: Cron Jobs, Scheduled Imports and Exports
• Automation Recipes for Ecommerce
• Case Study: Toy Brand Automated Feeds to 7 Marketplaces

Frequently Asked Questions

Does MicroPIM push product feeds directly to Google Merchant Center, or does it just generate export files?

Both delivery methods are supported. MicroPIM can generate feed files (CSV, XML, TSV) for manual or scheduled download, and it can push feeds directly to Google Merchant Center via the Content API, to Meta Commerce Manager via a feed URL or API, and to eMag via their product feed endpoint. Direct API delivery means you do not need to manage file hosting or configure channel-side feed URLs. The pet supplies company in this case study used direct API delivery for Google and Meta and file-based delivery for eMag.

How does MicroPIM handle product availability sync between WooCommerce and advertising channels?

MicroPIM connects to your WooCommerce store and reads current inventory status. When the export job runs, it reflects the current availability of each product at that moment. For daily scheduled exports, this means availability is accurate to within 24 hours. If you need near-real-time availability sync for channels where out-of-stock listings generate wasted ad spend, MicroPIM supports more frequent export schedules (hourly for channels that accept updates at that frequency).

Can we configure different product selections for different channels within the same feed automation setup?

Yes. Each channel template in MicroPIM includes an optional product filter — a rule set that determines which products are included in that channel’s export. Common uses include excluding certain product categories from Google Shopping, limiting the eMag feed to products that have a valid EAN code, or restricting the Meta feed to products above a minimum price threshold. The pet supplies company used this to exclude a product category from Google Shopping that consistently had policy issues, while keeping that category in the eMag and Meta feeds.

What happens to our feeds if MicroPIM experiences downtime during a scheduled export?

Scheduled export jobs that do not complete are retried automatically. The last successfully completed feed remains active on the channel during any gap. For businesses where feed staleness is a serious concern, MicroPIM’s job completion monitoring can alert your team if a scheduled job does not complete within its expected window.