Once we collect the data regarding offers of all the selected stores, the second stage of our work begins: matching your offers and those of your competitors, in order to compare the prices of the right products in the reports.
On average, PricePanorama catalogs about 150 thousand different offers for one client. It is a task impossible to be performed manually, by any human. We createalgorithms, that automate the cataloging for each type of assortment that we have to catalog. Algorithms analyze product names, their descriptions on store websites, and even photos to assess which products should be compared.
These algorithms vary greatly in the way they operate, depending on the industry.
In case of some algorithms, before implementing them, we need to create a catalog of the range of all products available in a given industry in order to 'teach' the algorithm the data that should be identified in product descriptions.
The most important factor affecting data quality is the creation of matching guidelines for products in each category with the client. This way we know how to program the algorithms and also what criteria we have to stick to when verifying data.
Our specialists verify the list of products compiled algorithms every day. They then validate the data provided, reject them or they consult on the next steps to be taken to achieve the assumed data quality with the programmers and our client.
The scope of such verification is determined each time in the contract with the client. For clients who automate price management, we offer a data quality SLA,which we achieve with a very strict data verification policy.
We use cookie files. If you would like to learn more about cookie files, the purposes for which they are used and the possibility to block them on your device please read our Privacy Policy..
Terms | GDPR for counterparties, users
Manubia Sp z o.o. © 2024