The challenge
Product data that held Google Shopping back.
MDM, the Münzhandelsgesellschaft Deutsche Münze, is one of the leading providers of collector coins and medals in Germany. The goal was to maximize product visibility and performance in Google Shopping, and the key to that is product-data quality.
A significant part of the product feed had optimization potential: products that generated no clicks (so-called zombie products) and items disapproved due to specific content.
Commemorative coins and bars were especially tricky, because Google Ads excludes advertising of means of exchange whose price depends on variable currency-market values. The task was to find precise, policy-compliant wording while making the ads more attractive.
Our approach
Product-data optimization in five steps.
Instead of turning bidding dials, we made the product data itself the object of optimization, AI-driven, policy-compliant and quality-assured. The steps:
- 01
Feed preparation
The existing product-data feed was analyzed and prepared for optimal processing by a large language model.
- 02
Prompt tuning
Based on client-specific requirements and proven wording, we tuned the prompts for high-quality, policy-compliant titles, descriptions and product highlights.
- 03
Automated optimization
adSoul optimized the selected products automatically: titles 77% longer, descriptions 41% longer, and problematic terms replaced with policy-compliant ones.
- 04
Quality assurance
A combination of automatic and manual checks secured data quality and allowed fast corrections where needed.
- 05
Go-live
The optimized data was published via a supplemental feed in Google Merchant Center.
- zombie products with no clicks in the feed
- items disapproved by Google over content
- short, low-relevance product titles
- sparse descriptions
- problematic terms for commemorative coins
- inactive products visible and relevant again
- policy-compliant wording, fewer disapprovals
- titles 77% longer and richer in detail
- descriptions 41% more comprehensive
- higher click and conversion rates
The results
More conversions, lower cost per conversion.
Comparing 60 days before and after the change, performance improved across all campaigns. The number of products without conversions fell by 24% to 48% depending on the campaign, and previously inactive products became visible and relevant again.
Products that already converted gained strongly: items with 10 or more conversions by 83% to 400%, products with 2 to 5 conversions by 33% to 68%. At the same time, cost per conversion fell by up to 9%, and the conversion rate rose by up to 17.5% depending on the campaign.
With adSoul we found a partner that not only understood our product-data management challenges but solved them. The results speak for themselves, we are thrilled with the performance increase.
What came next
The turbo for Google Shopping Ads.
The systematic, automated product-data optimization saves MDM a lot of time creating and maintaining its Shopping Ads and unlocks the full potential of the product catalog. On this foundation, optimization can be continued and expanded over time.
