Fiwe launches Inteller – identify ecommerce errors before your customers do

Traditional data quality checks establish whether information is present: is a colour specified? A size? A material? Yet a product can pass all these checks and still be described incorrectly. A T-shirt with “wood” listed as its material is a valid data point, but it is not valid information.

Similarly, an image might show matt tiles while the attribute says glossy, or a product might be recommended for both walls and floors while the description states that it is only suitable for walls. The fields are filled in and the data checks all pass, yet the information is still wrong. These errors are often discovered at the very end of the chain, by the customer, resulting in a support request, a return or manual correction work.

Fiwe is now launching Inteller, a service that reviews product information as a whole, rather than field by field. The service identifies contradictions, implausible values and other inconsistencies. Findings are grouped so that the person responsible for quality can see where the problems are and what needs to be addressed.

– Today, there is no process in place to identify incorrect information. If we are lucky, a customer may have contacted a salesperson to point out that a product image is wrong. Add growing volumes of data from different sources and an accelerating pace, and keeping up with both identifying and correcting errors becomes a real challenge. That is why we are pleased to launch Inteller: your tireless colleague, always ready to help you and your team improve information quality, says Mitra Javadzadeh, Head of Business Development at Fiwe.

The product catalogue is evaluated and the results are brought together in a single view. The person responsible for quality can then prioritise issues and track progress, while the owner of the product information makes the corrections.

– You can have one hundred per cent data quality and still have almost no information quality. With Inteller, we automate holistic validation and can start where it delivers the most value. Most organisations already have ticketing systems and backlogs for correcting errors, so what we add is detection. Corrections can continue to be made where they are today, in the systems people already use, says Martin Vollrathson, Developer at Fiwe.

The first version of Inteller is launching now, with development continuing on an ongoing basis.

Want to explore the difference between data quality and information quality? Listen to the Syntra episode “Can you trust your data?” or read our article on the topic.