Christian Hunger
Head of Category Management PCP.COM
Christian Hunger
Attributes mapped

+4800

Attributes mapped
Products categorized

+70.000

Products categorized
Suppliers processed

16

Suppliers processed

TOP CUSTOMER EXPERIENCES AND SEO-FRIENDLY PRODUCT PAGES

The PCP.COM Group has been active in online retailing since 1998. As a pioneer and leading Swiss multichannel provider for computers and home electronics, the PCP.COM Group operates 14 specialty stores in addition to several online stores in the DACH region with more than 700,000 items from over 5,000 manufacturers and around 250 leading suppliers. The group includes, for example, the well-known brand STEG Electronics or the technology online store techmania.ch.

The market is extremely competitive, particularly in the IT and consumer electronics sectors, which are dominated by major international players and high pressure on margins. To survive in this market, PCP.COM relies on a broad product range with high-quality product data and informative content. To achieve this goal, PCP.COM has chosen Onedot. This is because Onedot is the only way to automatically create high-quality product data that is customer- and search engine-friendly.

HOW ONEDOT HELPED

For PCP.COM, the focus is not only on expanding the product range for its online stores, but also on improving product data quality in a timely manner in order to offer even better customer experiences and thus increase conversion. After all, in the highly competitive electronics market in particular, a very good customer experience is a key factor in the purchase decision. This means that only enriched and high-quality product descriptions also lead to sales success. In order to achieve this goal, a streamlined and automated data preparation process was introduced with Onedot as a basis.

This data preparation process was first introduced along clearly structured steps. First, the product data model of PCP.COM was initially mapped in the Onedot platform. This means that the category tree, all attributes and data types, as well as formatting patterns etc. were created on the platform. Subsequently, the existing product master of PCP.COM was used to train Onedot's software and associated PCP.COM-specific machine learning models. This preparatory step was central to triggering automated product data onboarding.  

Once the product data platform had been set up and trained, the onboarding of individual suppliers could begin. Thus, supplier after supplier and product catalog after product catalog were loaded onto the product data platform and then the onboardings were initiated. The product data in various forms and classifications was then transferred to the target data structure of PCP.COM. Since some suppliers do not provide sufficiently high-quality product data, the content provider icecat was also connected via Onedot. In this way, PCP.COM uses additional content from content providers in addition to the content from suppliers/distributors in order to create golden records with maximum information content. In this way, PCP.COM achieves better product data with a high degree of up-to-dateness compared to other retailers.

Regardless of whether the product data is supplier data or product data from a content provider, product data goes through the same Onedot structured onboarding process. Each of the onboarding steps can be viewed, checked and validated by PCP.COM thanks to so-called feedback tasks. Thus, in the first step, product data from suppliers and from the content provider is assigned to PCP.COM's categories and each of these category suggestions is presented with a confidence level (low, medium, high) as a basis for validation. The higher the confidence level, the more likely it is that the product has been assigned to the correct category. Furthermore, validated mappings are used in the next catalog update without the need to check them again.

Subsequently, the attributes of the suppliers are assigned to the attributes of PCP.COM. This is done based on word similarities, statistical properties and semantic methods by Onedot's software, which then automatically suggests mappings with the appropriate confidence level. Again, users can view these suggestions and override them if necessary. Validated mappings do not have to be checked again in this step, provided they have already been validated in onboardings that have already been carried out. The feedback received is used by the software to constantly update the supplier profile or the data of the content provider as well as the machine learning models. This leads to increasingly accurate suggestions for new mappings and even less checking effort by the PCP.COM data team.

Finally, the normalization step takes place, in which numbers and units as well as value lists are normalized and converted. Here, too, suggestions are generated with confidence levels that can be viewed and overridden by the user if necessary. Thanks to this automated data preparation process, not only is the quality of the product data significantly increased, but the data team at PCP.COM is also relieved. The data team only has to monitor the overall process and can devote itself to the strategic orientation of the online store and other important issues.

THE RESULT

Since January 2022, 112 onboardings with more than 600,000 SKUs have been carried out via the Onedot product data platform. The update of the PCP.COM product data model in the Onedot product data platform also runs in the background so that changes in the PCP.COM data model can also be taken into account in the Onedot platform.

Thanks to Onedot, the broad product range can be further expanded quickly and with high quality. Because many processes are automated, the products are always up to date and available in the best e-commerce quality.

It was the right decision to buy the onboarding solution from Onedot and not to build it ourselves.
Martin core
Martin Kern
Head of IT and Development
Martin Kern
Head of IT and Development
Christian Hunger
Wir sind einer der grössten unabhängigen Detailhändler der Schweiz

See for yourself.

In order to be able to expand our online product range quickly and in a high-quality manner, we rely on the product data platform of Onedot. In this way, we have digitalised our data maintenance processes and found a strategically important partner in Onedot, which also brings relevant machine learning expertise.

Read the
success story.

Replacing our existing IT applications with a future-oriented e-commerce infrastructure is a strategic goal for us. Onedot convinced us as a product data onboarding solution, as a high degree of automation can be achieved via a modern AI when connecting suppliers and data platforms.

Read the
success story.

Through Onedot we can ensure efficient supplier onboarding and accelerate the expansion of our marketplace. In doing so, we achieve a new level of digitalization in internal processes.

Read the
success story.
Experience Onedot with your own eyes.
See for yourself how much high quality product data can bring to your business. Free of charge and without obligation.
Schedule a demo