20 million records a day: how Azerty keeps product data moving at scale.
One central source. Automated processing. Live within 10 minutes.
Azerty, part of the EyeTi Group, is a Dutch online retailer and IT specialist managing more than 400,000 SKUs across hardware, components and business IT products.
That range is growing fast and is sold through multiple webshops and marketplaces.
With around 100 connections to suppliers, distributors, manufacturers and data sources such as Onetrail and Icecat, Azerty processes roughly 20 million records every day.
At that scale, product data quickly becomes an operational bottleneck. Not because PIM falls short, but because supplier data arrives in a continuous stream: new supplier data and updates to existing product data, which constantly need to be processed automatically.
ConnectingTheDots helps Azerty structure that incoming data and pass it automatically into one central product data flow in PIM.
- 20 million: records processed automatically every day
- 400,000+: SKUs managed centrally
- < 10 minutes: from new product to live publication
The challenge: product data at operational scale.
Azerty works with a large number of suppliers, manufacturers, distributors and data sources.
Each party supplies data in its own way. Think of different structures, price lists, product attributes and update frequencies.
At low volumes, teams can absorb much of that variation manually.
At roughly 100 integrations and millions of records per day, that approach no longer scales.
The challenge was therefore not only to manage product information centrally, but above all to make sure supplier data is onboarded and processed in PIM automatically. At scale.
The acceleration happens upstream of PIM.
A key part of the PIM platform setup is Import & Onboarding.
Supplier files and price lists are ingested automatically, analysed and mapped to the structure Azerty needs.
This means a large part of the recurring preparation work is already done automatically before PIM.
Teams no longer need to process these files manually. Their attention can shift to exceptions and situations that require judgement.
Fragmented supplier data comes in. Structured product data flows automatically from PIM to the sales channels.
Read more about Supplier Data Onboarding.
400,000+ SKUs in one central product source.
The full range of more than 400,000 SKUs is managed centrally in ConnectingTheDots PIM.
Automatic sales price calculation is also part of the product data flow in PIM.
From that central product source, Azerty supplies:
- two webshops;
- its ERP system;
- a data lake;
- four marketplaces.
In PIM, product data is structured automatically once and can then flow in a controlled way to different destinations, via Publishing & Channels and Integrations.
Product data is structured once and can then be distributed in a controlled way to the systems and channels that need it, through Publishing & Channels and Integrations.
From supplier data to live in under 10 minutes.
Speed does not come from publishing faster in isolation.
The real gain comes from removing manual steps between supplier input and downstream publication.
When a new product enters the process, data can move automatically through:
- import;
- classification;
- validation;
- enrichment;
- central storage;
- distribution to connected systems and channels.
For Azerty, that means a new product can be live within 10 minutes.
Teams focus on exceptions.
At this scale, it is not realistic to check every product manually.
The setup is therefore fully focused on processing product data automatically.
Teams step in when data is incomplete, inconsistent or requires additional review.
That changes the operating model.
Instead of pushing millions of records through the chain by hand, the team can focus on exceptions and on improving the rules that drive the automated product data flow in PIM.
Azerty shows what product data automation delivers: more volume does not have to mean more manual work.
The result.
Azerty has one fully automated product data flow in PIM for a range that keeps growing and keeps changing.
Azerty's case shows that product data automation at scale makes it possible to process millions of records and hundreds of thousands of SKUs continuously, without more volume automatically meaning more manual work.
- around 20 million records processed automatically every day;
- 400,000+ SKUs managed centrally;
- around 100 supplier and data-source connections;
- new products live within 10 minutes;
- one central source serving webshops, ERP, a data lake and marketplaces;
- teams focusing on exceptions rather than repetitive processing.
For Azerty, PIM is not a standalone system beside the operation. It is part of a product data architecture that connects supplier input, central product management and downstream channels.