GCPProductionEvidence: Medium65/100

Thomann.io replaces fragmented recommendation system with Vertex AI Recommendations

Use case typeCustomer targetingUpdated Jun 13, 2026

Thomann.io replaced a 13-year-old fragmented recommendation system with Vertex AI Recommendations to improve personalization across newsletters, mobile app, and website. The implementation used BigQuery, Bigtable, Google Analytics, Cloud Composer, and partner support from adesso.

Organization: Thomann.io
Industry: Retail
Location: Germany
Published: May 2026

Reported outcomes

Revenue: +3%

Revenue & growth

Impact: +13%Revenue: +24%

Catalog median for revenue & growth deployments: +40% across 67 reported metrics. Compare benchmarks →

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Quantified impact: 3% increase

Google Cloud Customer StoryMay 22, 2026Customer storyInferred claimMedium evidence strength

3% increase in conversion rates across customer touchpoints

Normalized claim

Quantified impact: 13% increase

Google Cloud Customer StoryMay 22, 2026Customer storyInferred claimMedium evidence strength

13% more items added to cart when recommendations were shown

Normalized claim

Quantified impact: 24% increase

Google Cloud Customer StoryMay 22, 2026Customer storyInferred claimMedium evidence strength

24% increase in average order value

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Thomann.io, Thomann
Provider: GCP
Maturity: Production
Linked source: Google Cloud Customer Story

A fragmented rule-based recommendation system struggled with an extremely diverse product catalog Personalization logic could not handle varied customer journeys and technical debt was increasing Migrated from rule-based logic to Vertex AI Recommendations Deployed recommendations across newsletters, the mobile app, and website Used BigQuery, Bigtable, and Google Analytics to process product and behavioral data Implemented A/B testing rollout and business rules for shipping constraints, consent f

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Personalization
2Recommendation engine modernization
3Customer experience

Migrated from rule-based logic to Vertex AI Recommendations
Deployed recommendations across newsletters, the mobile app, and website
Used BigQuery, Bigtable, and Google Analytics to process product and behavioral data
Implemented A/B testing rollout and business rules for shipping constraints, consent fallbacks, and fairness controls

Technologies

Vertex AI Recommendations BigQuery Bigtable Google Analytics Cloud Composer

3% increase in conversion rates across customer touchpoints
13% more items added to cart when recommendations were shown
24% increase in average order value
High two-digit ROI within the first year

Architecture

Thomann.io migrated from a rule-based recommendation engine to Vertex AI Recommendations on Google Cloud, using BigQuery for catalog and customer data, Bigtable for real-time session and email relationship data, and Google Analytics for behavioral signals. The rollout progressed from newsletters to app to website with A/B testing and business-rule controls for consent and shipping restrictions.

Sources & evidence1

Evidence: Medium65/100Evidence strength