Sebastian
How DOUGLAS Tames 16 Million Products with AI – Without Formula Chaos
Imagine having to manually format 16 million products - creating titles, assigning categories, correcting errors. Challenges like these were the reason why DOUGLAS had to completely rethink product data optimization. The solution? Artificial intelligence, which translates complex data structures into understandable language, saving time, resources, and stress. The company uses our AI Formula Generator in Channel Pilot Pro and impressively demonstrates how even huge product catalogs can be optimized cleanly, quickly, and platform-appropriately. Without script confusion. External agencies? Superfluous. Formula chaos: History.
DOUGLAS: Premium Beauty at Enterprise Level
DOUGLAS It's not just your average local perfumery. As Europe's leading premium beauty provider, the company can boast impressive figures:
- 1,900 Stores throughout Europe
- 38 Online Stores in 26 countries
- A Centralized Performance Marketing Team in Düsseldorf
- A Monthly SEA budget in the millions
This means: Extreme advertising pressure, high responsibility - and zero tolerance for data chaos. In our Live session at the K5 Conference 2025 In Berlin, together with Jonas Franke, Team Lead SEA at DOUGLAS, we presented the new way of product data management: Powered by AI. No code. Faster than ever before.
„Each SEA manager handles three to four countries simultaneously. And this, of course, presents us with huge challenges, especially with regard to automation solutions that we urgently need.“
Jonas Franke
Team Lead SEA, DOUGLAS
Why product titles determine success in online retail
Product titles are often the first thing users see – whether on Google Shopping, Idealo, or in the Douglas app. They determine:
- Visibility in the search results
- Click-through rate and user engagement
- Cost per click (CPC)
- And ultimately over the Revenue
The Challenge: Every platform wants its own syntax. Google likes it clean. Criteo is a bit more lenient. Faulty feeds? Campaign blocked. Time for an intelligent solution.
The Formula Logic Nightmare: When Code Becomes the Problem
For years, DOUGLAS worked with manual formulas, creating a complex web of nested conditions that, while functional, became increasingly verbose. Furthermore, this approach posed the risk of creating knowledge silos. This means that only a few colleagues possess the know-how to efficiently manage product data.
The typical challenges:
- Complexity grows with every special case
- Formulas become unmaintainable and error-prone.
- Changes take hours
- Debugging becomes a full-time job
„A small mistake could cost us tens of thousands of euros, which we would lose.“
Jonas Franke
Team Lead SEA, DOUGLAS
The Solution: The AI Formula Generator
With the smart AI Formula Generator in Channel Pilot Pro, formula work becomes a casual conversation. At DOUGLAS Product Data Optimisation Smart technology is used to translate complex data structures into understandable language:
💬 Simple prompt „Create a product title that meets the following criteria: Combine brand, product line, product name, category, product addition, size, color, and gender.“
Done. No code needed. Debugging? Superfluous. And frustration? Out of the question.
Here's how it works in practice
DOUGLAS uses AI to optimize over 16 million products. Some typical use cases include:
Common Feed Issues:
- Double size indications
- Wrong language („Female“ instead of „Ladies“)
- Unnecessary details (e.g., „unisex“ on hand sanitizer)
- Formatting error (e.g., „50.0 ml“ instead of „50 ml“)
AI-powered solutions
- Cleaning with „If-Then“ Rules in Plain Language
- Automatic language and format corrections
- Filter logic controllable directly via prompts
- Scalable rules for millions of products
You can find more use cases in our live session.
„It will actually be suddenly usable for everyone because the hurdle is just very, very small.“
Carsten Mesecke-von Rheinbaben
CCO, Channel Pilot Solutions
Product categories made easy
In addition to product optimization, DOUGLAS AI is also used for intelligent campaign logic. Example: Product „pots“ for strategic campaign control. By dividing articles into groups, they can be targeted later. This is comparable to a filter. DOUGLAS itself calls the groups "pots".
Every pot = individual bidding strategy for maximum performance.
- Top A: Private label with high margins
- Top B: Products with a profit margin of >300%
- Top C: Mid-range products with a margin of >220%
- Top D: Exclusive brands
- Top E: Everyone else
The best part: Define rules in understandable language and have them converted into working code by AI.
The advantages at a glance
| Advantage | Effect |
|---|---|
| AI instead of Formula Code | No syntax errors, no maintenance hell |
| Fast iteration | Optimizations in minutes, not days |
| Platform-appropriate & scalable | Works for 500 or 5 million products |
| Less agency dependence | Expertise remains within your own team. |
| Clean Data = Better Campaigns | Higher visibility, lower CPCs |
How to: Product Data Optimization with AI
Ready to transform your product data management? Then you should go through this little checklist:
Step 1: Identify pain points
- Where do you spend most of your time with manual formatting?
- Which platforms require different data structures?
- What are the most common mistakes?
Step 2: Start with Simple Prompts
- Start with basic „Use attribute X, exclude Y“ instructions.
- First, test with small product sets
- Iterate based on the results
Step 3: Scale Gradually
- Apply successful patterns to larger datasets
- Create templates for frequent scenarios
- Expand your prompt library continuously
The future is digital conversational
DOUGLAS proves that modern product data management works: Fast. Scalable. Understandable. And above all: Accessible to everyone.
Whether for product titles, pricing logic, or categorization – AI in Channel Pilot Pro eliminates hours of Excel clicking and formula debugging.
Sebastian
Lead Design & Event
At Channel Pilot Solutions, Sebastian is responsible for Design and Events – with the goal of presenting complex topics in a way that makes them understandable and memorable. From his many years of experience as an Art Director, he brings a keen sense for brand impact, clear design, and details that make the difference in the end.
In his contributions, he demonstrates how sophisticated channel pilot topics can be made visually and emotionally tangible – and why good design not only looks good but also measurably strengthens communication. He finds the necessary balance in creative and mindful moments away from everyday work.
Deutsch 
English 
Español 
Français