Glow25 is a Berlin pro-aging brand for collagen-based beauty and wellness products, sold online across Europe. Gemma Analytics built Glow25 a modern data infrastructure on BigQuery, dbt and Lightdash, with self-service dashboards and an AI interface for questions in natural language.
The starting point
Glow25 is online-first. It sells through its own webshops on Shopify and WooCommerce, through European marketplaces such as Amazon, eBay, Kaufland and Shop Apotheke, and through subscriptions with fast shipping. Every channel delivers data in its own format and with its own logic, from the Reybex ERP to the marketing and CRM tools.
Fragmented, largely manual data processes made it hard to track key business metrics. KPIs such as customer acquisition cost (CAC), channel conversion and churn could not be monitored efficiently, and non-technical teams had no quick way to answer their own business questions. Fast, independent decisions were out of reach.
What we built
Gemma Analytics started with a proof of concept for a central performance dashboard and built the full data platform from there. All e-commerce, marketplace, marketing and subscription data now flows into Google BigQuery as one single source of truth, loaded by Apache Airflow and Fivetran. Transformations moved from Dataform to dbt: dbt models turn raw data into clean, tested metrics, and every sales channel lands in one shared order model with a revenue date that fits the channel.
For reporting, Glow25 moved from Metabase to Lightdash. Fast, maintainable dashboards for e-commerce, subscription and P&L performance refresh daily and give every team direct access to curated numbers. On top sits a semantic layer with an AI interface: teams ask questions in plain words and get answers from the same governed definitions the dashboards use. Customer email addresses are pseudonymised early in the pipeline, and automated tests flag missing product costs before they distort the margin figures.
The result
Glow25 now has one foundation for e-commerce, subscription and P&L reporting. KPIs like CAC, churn and conversion update automatically every day instead of being compiled by hand, and business teams find answers themselves in Lightdash or simply ask in natural language. Glow25's own data team builds further models and dashboards on the same semantic layer, so the platform grows with the company.
Tools and setup
Sources include Shopify, WooCommerce, Amazon, eBay, Kaufland, Shop Apotheke, the Reybex ERP, Meta, Google Ads, Google Analytics, Klaviyo and Braze. Apache Airflow and Fivetran load the data into Google BigQuery, the cloud data warehouse. dbt handles modelling, testing and documentation, and Lightdash provides self-service BI for business teams and management, together with the AI interface on the semantic layer.