How we used Event Storming to improve a PIM process

Nikolaus Varzakakos
March 27, 2022
2 min

In a recent project, we facilitated several Event Storming workshops for a successful e-commerce company to improve its Product Information Management (PIM) process. The client manages nearly a million SKUs and frequently updates them from a diverse group of suppliers. Our goals were to:

  • Decrease time to market for new products
  • Improve the quality of product information
  • Improve internal work processes

The Challenge

Before new products can be published in the online store, product information must be captured, enriched, and reviewed in multiple process steps such as:

  • Retrieving product data from suppliers
  • Verifying the data against minimum requirements
  • Transforming it into a common structure
  • Enriching it with images, videos, marketing copy, size charts, supplier information, etc.

To reduce time to market, we looked for opportunities to increase automation. It was crucial to engage the teams responsible for each step. Some of the key roles included:

  • Data Manager
  • Content Integration Specialists
  • Content Production Specialists

Our Approach

Event Storming is an ideal method for tackling complex business processes involving multiple teams and steps. It brings domain experts and engineers together to build shared understanding and align on improvements.

We planned a series of Event Storming workshops with around 10 participants from various departments. Using Qlerify, we started with a blank canvas prepared with standard Event Storming notation: card types for Problems/Hot Spots, Opportunities, and Systems, plus swimlanes for roles.

As we've experienced many times before, Event Storming quickly engaged the participants. The discussions were productive and focused as we collaboratively mapped out all events, roles, and challenges in real-time.

The Impact

After just two hours, we had outlined the main flow and started digging into sub-flows. By the third hour, we had already reached consensus on the most important problems to prioritize.

Previously, reaching this level of clarity could have taken weeks—if it happened at all. Thanks to Event Storming and Qlerify, we achieved it in just a few working days, saving possibly more than 80% of the time.

An added bonus: communication between departments improved significantly. Before using Event Storming, people often talked past each other, each with their own mental model of the process. Event Storming helped create a shared understanding and alignment around priorities.

Want to learn more about Event Storming or Qlerify?

Visual workflow of a Product Information Management (PIM) process using Event Storming. The diagram shows automated and manual steps across three lanes: Automation, Data Admin, and Marketing. It starts with scenarios for retrieving product data via API, file uploads, or external sources. Data is automatically transformed and verified, followed by manual steps like data cleaning, category updates, and enrichment. Products are then approved and published across web, marketplace, self-service portals, and social media. The process ends with product maintenance and end-of-life removal.

‍

Recent posts
One order passes through five systems with Excel and email in between, compared with resource thinking and flow thinking.
Enterprise AI: Running AI Across Teams and Systems
Enterprise AI runs a whole process across teams and systems. These are the six things to put in place first, and how long a typical rollout takes.
Staffan Palopää
Oct 29
7 min
Spec-driven development still leaves out the people who know the business
Spec-driven development fixed vibecoding for developers. The people who know the business still can't review the spec. Here's how Qlerify puts them in it.
Nikolaus Varzakakos
Oct 14
5 min
AI Legacy Modernization: Reverse-Engineer Legacy Code into a Domain Model
Use AI to reverse-engineer an enterprise-grade business application, adapt it to your needs with humans in the loop, and reimplement it on your selected tech stack.
Staffan Palopää
Sep 17
10 min