Galeria: From vision to measurable performance with Google Cloud AI

In brief

How does a traditional retail company make the leap to an agile, AI-supported retail brand? Together with e-dialog, Galeria shows how the use of Google Cloud AI not only modernizes the shopping experience, but also proves the business case for innovation directly through measurable performance increases. The goal: a highly efficient, personalized digital journey that creates real business value.

Personalization as a sales driver – validation of Vertex AI Search

Business Case 1

Personalization as a sales driver – validation of Vertex AI Search
Galeria proves through a scientific test setup by e-dialog that AI-supported personalization in search achieves double-digit sales increases and thus makes the technological transformation directly profitable.

Challenge:

The previous search was at its limit. The switch to Vertex AI Search meant a switch to usage-based costs – Galeria therefore needed statistical proof of the commercial upside before the rollout.

Measures:

e-dialog established a precise validation framework:

  • Data Hygiene: Optimization of product feeds for model training.
  • Scientific testing: Large-scale A/B test (1.5 million sessions).
  • Precision Tracking: Granular allocation of every euro of sales to the AI result.

Results:

The test proved that AI relevance pays off. Despite stable click rates, purchase intent increased massively.

  • % Revenue Rate Lift
  • % Purchase Rate (Conversion) Lift
  • % Add-to-Cart Lift

We wanted to shape the future of search, but we had to be sure that the business case would hold up. e-dialog helped us to validate this. It wasn’t the promises of AI that convinced us, but the figures in the test. The 15% increase in sales proved that the move into this new technology was exactly right.

Bert Middendorp
eCommerce Director, Galeria

Further information on GenAI & Cloud for eCommerce cases


 

A scalable video engine through generative AI

Business Case 2

A scalable video engine through generative AI
Galeria scales content production to up to 250,000 videos in three months using an AI engine developed by e-dialog, reducing costs and time by 90%.

Challenge:

Product videos are traditionally too expensive and time-consuming for the entire catalog. Galeria was looking for a solution to create high-quality video content efficiently and automatically on a large scale.

Measures:

Use of Google Vertex AI VEO for automation:

  • Production: Creation of 180° orbit shots and AI-generated scenes.
  • Infrastructure: e-dialog developed a high-performance architecture on the GCP.
  • Quality: Integration of an automated module to ensure brand consistency.

Results:

Galeria established a video engine that radically minimizes costs and time.

  • Product videos produced in 3 months
  • % Brand Consistency through AI quality check
  • % Cost savings compared to traditional production

The integration of VEO allows us to overcome the traditional time and cost hurdles of studio video production. By creating high-quality product videos on a large scale, we are not only further developing the digital shopping experience, but also positioning Galeria as a more agile, AI-driven brand in a competitive market.

Bert Middendorp
Director of eCommerce, Galeria

Conclusion: Truth in Numbers – Why holistic AI strategies win

The cases show: Innovation does not need advance praise, but validation. While the personalization of search (Case 1) creates the necessary relevance and secures the budget for further growth through direct sales increases, the Gen AI Video Engine (Case 2) enables the efficient scaling of emotional content.

Key takeaways for your business:

  1. 01

    Performance finances progress

    Use clean A/B tests to prove the ROI of new technologies – turning AI from a cost factor into a return driver.

  2. 02

    Scalability through infrastructure

    A smart cloud setup (GCP) can handle massive amounts of content without producing exploding costs.

  3. 03

    Customer experience as a goal

    Whether more precise searches or dynamic videos – in the end, the brand that offers users the most relevant experiences wins.

Photo credit for portrait photos of Bert Middendorp: StudioF_YvonneFetz

Know-how

Exciting news & relevant updates