background Layer 1

PIM - product catalog management system


Georgy Khotyan , head of development of back-office systems at Detsky Mir, talks about how PIM was created, its internal structure and main advantages .

The PIM (Product Information Management system) system is designed for storing and processing large volumes of content data about a product:

  • main ones – name, article, barcode;
  • additional – specific to each category, of which there can be thousands. This could be color, material, fabric composition of clothing, etc.;
  • media data – photos, videos, documents, instructions;
  • logistics – weight, size and so on.

How the system came about

We started writing the program from scratch, and the previous solution worked for 13 years. When it no longer met the requirements, we decided to rethink the approach. Let us explain why we had to do this.

Why did you decide to develop your own PIM product:

  • The solution available in DM is technologically outdated (it worked for 13 years);
  • The product used in DM had foreign owners;
  • With an increase in the assortment in the DM online store, unstable operation of the vendor’s system was observed;

Before implementing the project, we studied a number of existing ready-made solutions, but decided not to dwell on them. The main problem was the need for strong customization and integration with our current systems. The main task was outlined - to integrate PIM without significantly changing them.

We work with several trading platforms. Their list is presented in the figure:

At all Detsky Mir sites (Kazakhstan, Russia, Belarus), the assortment is the same, but a different approach to publication is required. For example, a product may only be sold in one country. Zoosaurus has a different assortment than Detsky Mir, but it overlaps with the Yet discounter. To some extent, Detsky Mir's products also overlap with the discounter. It is more convenient for the content department to work with the assortment without thinking about where it will be sold, but at the same time, data about the products should go to the appropriate sites.

There is a separate supplier portal (personal account), with which there should also be close integration. We have more than 4,000 suppliers, so this is a large amount of information. One of the main goals of PIM is to give them the opportunity to enter all the characteristics of the product themselves, without further processing by our content department. Now they fill out a set of data that is the same for all categories. The idea is to pass a dynamic catalog system to the supplier to populate. This will simplify the work of content specialists and speed up data processing.

What is the system responsible for?

PIM is the data system on which the storefront is built. Therefore, it’s worth telling more about the storefront and what PIM is responsible for:
  • search;
  • category tree (in the figure under number 2);
  • sorting – by popularity (default), novelty and other criteria;
  • media data;
  • dynamic filters are actually a PIM innovation. For each category, the order of the filters themselves and the order of the attributes within the filter are dynamically adjusted. This is a capability that the previous system did not provide;
  • Card Product.

In the product card PIM is responsible for:

1. Well-known “breadcrumbs” – a product can be included in different categories, but the system forms all the data about it into one main category.

2. Product name. Suppliers enter their names for one product, and PIM, according to the template, generates a uniform name (for example, clothing will necessarily have a size and color).

3. Media data.

4. Brand with the corresponding logo (comes in a separate directory).

5. Product characteristics.

All this increases conversion. The ability to view information in a structured, sorted form and filter out what the visitor needs encourages them to make a purchase.

Internal structure

We have taken the path of simplifying the internal category system, which is convenient for suppliers to work with. For example, within PIM there is a single category “Children’s clothing”, which on the site can be transformed into anything (for example, “Dresses for girls”, “Children’s New Year’s costumes”, etc.).


Each category has a PIM information model attached to it, and the attributes are linked to the information model. For example, a set of attributes of the “Toy Transport” information model can inherit many categories – “Cars”, “Airplanes” and others. This inheritance gives greater flexibility when working with categories. The creation of information models contributed to achieving the first goal – accelerating the filling of product content.

Front category management

The next goal was to quickly create categories and category trees.

Site hierarchy - creation of hierarchy for external systems (Detsky Mir site, More site, Zoosaur site). 

Mapping rules – criteria for automatic distribution of assortment into hierarchy, based on product card data.

Listing view – customisation of the listing view on a particular site, display of filters in a category

Product presentation – customisation of product presentation on a specific site, output of product properties block.

Boundary management – product presentation (display) on a specific site depending on commercial criteria.

We followed the path of a generalized structure, which, using mapping rules, is transformed into its own category trees for “Children’s World”, “Zoosaurus”, and a discount store. For example, “Children’s clothing” by age group can be transformed into “Clothing for babies”, and by season – into “Summer clothing”. Categories on websites are now filled in automatically; employees working with content don’t think about it.

Card management

Creation of Master Cards (MC) – product card with properties based on binding to Master Category and infomodel

Creation of partner cards (SKU) – offer of a particular partner, with its unique content data

Editing fields of SKU – possibility to edit fields both through web interface and with the help of mass updates via Excel files Media uploading - possibility to upload photos, videos, documents to the master card, both individually and by mass uploading.

MC grouping – ability to group MCs by one common feature (e.g. colour: T-shirt – red, T-shirt – white)

Publishing – ability to publish MCs and SKUs on a specific site, in a specific country

In marketplaces, several suppliers often offer the same product. We have implemented the ability to work with one product by linking partner products to the main product card.

Administrative panel

MS view customisation - ability to build a table of MSs from any properties and in any sequence

MS filtering – the ability to filter the MS list by any property

Excel data upload - ability to upload all filtered MSs to Excel in a pre-configured format (same as the view setting)

MS upload by category - upload filtered MS to several files in the category model date view.
(1 file = 1 category)

PIM provides a panel with easy view customization. The ability to work with Excel files facilitates fast processing. If you need to add 10,000 products, add an attribute, we customize the listing, select the necessary parameters, upload it to Excel, make changes, and upload it back. This is a separate example of performing bulk operations.

Bulk Operations

  • Transfer of MC by category
  • Publish/Unpublish
  • Edit MS and SKUs via Excel templates
  • Upload photos
  • Upload files (instructions, certificates, etc.)

Mass operations – functionality that speeds up and facilitates work. This may be multiple publications or removal from publication on individual or all platforms. Photos are also uploaded through bulk operations. Images are selected from a photo archive; it is enough to apply a certain naming template. Files are downloaded in the same way.

Other functionality

  • History of changes – allows you to track the amount of work of external contractors, which is necessary for calculations.
  1. MS Change History
  2. SKU change history
  3. Category front history
  4. Master category change history
  5. Infomodels and properties change history
  • System analytics
  1. Integration error reports
  2. Interactive report on product creation
  3. Customisable reports for content manager tasks
  • Role model
  1. Supermoderator - can make any changes to the PIM
  2. Content moderator - can only work with MCs
  3. Viewing role - cannot edit anything, only view
  • Интеграция со всеми системами "Детского мира" - основной момент при разработке новой системы. Так, интеграция с медиасервисом позволяет быстро представить покупателю фотографию товара еще до загрузки текстового описания на сайт.
  1. Integration with Supplier's LC (Marketplace);
  2. Integration with SAP ERP; SAP BW;
  3. Integration with DM media service;
  4. Realisation of offloads for work with websites. 


Let us briefly list the features of the system:

Auto-naming - the ability to type product names, depending on the category according to certain rules

Mapping rules - the assortment is distributed in the hierarchies of the site according to predefined criteria and on the basis of product data

Mass operations - wide possibilities of various mass operations, significantly reducing manual labour

History - any changes are recorded. It is possible to unload the history for a certain object or for a user, for any range of time.

Category front - possibility to have different sites with different represented assortment and different hierarchy from a single product base.

Property restrictions in IM and Category Wizard with regular expressions - it is possible to set restrictions on entering both specific characters and values.

Property mapping builder for external category - possibility to convert PIM model date to another system's model date.

% of filled goods - calculation of filled properties in the goods both by infomodel and on the basis of the front category the goods fall into. 

This gives the following beneficial effects:
  • Reduce time spent on product data management
  • Improved accuracy of product information
  • Increase sales through improved customer service
  • Increased ability to analyse product data
  • Increased ability to integrate with other systems
  • Increased ability to provide real-time product information to customers

Questions and answers

How long does it take to index a catalog for display to users?

When sending the entire data array (processing a million products) using SOLAR, indexing takes place in an hour. If you only send changes daily, there will be hundreds of times fewer positions, and indexing will speed up significantly – it will take approximately up to five minutes.

Was there any point in writing a product from scratch?

We are glad we made this decision as the process is still in progress. Now we are developing integrations, the same team is doing this.

How many specialists were involved in writing and supporting the product?

Our team consists of three backend and two frontend developers, two testers, and an analyst. Several times we hired outside developers for short-term work.

The integration of Artificial Intelligence (AI) into business operations marks a transformative era, enhancing efficiency and innovation across industries. From revolutionizing HR with automated recruitment to aiding early disease detection in healthcare, AI's impact is profound. It enables predictive cybersecurity, personalized customer interactions, and accelerated software development in IT. 

Lots of news from technology vendors and modern cases on how to use data analytics for operations excellence – this is what March brought us this year.

In this issue we looked at the drivers and bright trends in retail. AI will undoubtedly be the leader among technologies for several years ahead in retail, and in the economy in general. Its emergence in wide access has opened up incredible opportunities for improving business efficiency.

“Before the automation introduction, the process was complicated by the fact that the address system was stationary, not dynamic. We could not automate serial and batch accounting of products. It was not possible to automate the warehouse replenishment processes, both planned and on order. The accounting system did not reflect all goods movements,” says Alexander Permyakov, head of the warehouse complex.

We use cookies for analytical purposes and to deliver you the best experience with our website. Continuing to the site, you agree to the Cookie Policy.