Automating data enrichment: why should your business start now?

Wholesalers and retailers are required to process a lot of product information, as they receive encrypted product data from suppliers, which then need to be transformed into qualitative product descriptions with columns of product features. The reason for this is that customers need to be sufficiently informed in order to decide whether or not to buy a product. When not sufficiently informed, customers may decide to buy from another store or lose their interest in the product at all.

Organizations that sell many products are confronted with the huge loads of data that need to be processed in a proper way. Within these loads of data there may consist some inconsistencies or mistakes that need to be fixed. Manually fixing these mistakes is almost unfeasible, especially for organizations with so many products. This article will take a closer look at how to fix inconsistencies in loads of product data. Moreover, it is discussed how to enrich product data and how this process could be (partly) automated.

Many suppliers deliver product data in different forms. This often results in encrypted data which need to be transformed: manually transforming these data to text is time-consuming, especially if you need to do it for thousands of products. This is, however, an important aspect of enriching data and can have significant impact on other business processes. Many master data managers still manually enrich data, which is not the most efficient method.

Automating the process of enriching data

In order to optimize the efficiency of the process of enriching data, it could be beneficial to automate it. This could save your organization much valuable time and many unnecessary costs. Data enrichment software nowadays can recognize encrypted information and directly transform this into textual information. is such an enrichment software. It uses all available product data and transforms the inconsistent or incorrect features into textual features. For unfilled data fields, the software is able to fill in the correct feature. For example, when a list of products is incomplete because the color of some products is missing, the software checks all available data on these products to fill in the colors. In this way, between 80-90 percent of all available information can be extracted and turned into tangible features. The smart tool comes with 4 functions which will be explained briefly.

Text extraction

The most frequently used function of this tool is the text extraction function. This function focusses on existing text sources with encrypted information: namely it extracts product features extracted from brand names, product names, product descriptions, feature lists or information from ERP systems. The algorithms of are trained with huge amounts of data and are able to learn from data models and for example recognize abbreviations and synonyms. After doing the hard work for you, the smart tool provides you with a list of proposed product features, which can then be manually validated. A screenshot of the text extraction function is provided below.

Screenshot of (text extraction function).

Image extraction

The second function of the software is image extraction. The smart tool is able to generate textual information based on images. The tool will extract data from the visible features, such as color, fabric or number of legs (in case of a chair) from the image. If there is text present in the image of for example a jar of peanut butter, the tool is able to read the texts and extract in this case all ingredients and nutrients. A screenshot of the image extraction function is provided below.

Screenshot of (image extraction function).

PDF scraping

The next solution of is the ability to extract data from PDF files. This can be useful for organizations that need to process PDF files in bulk. Within a short period of time, many PDF files can be scraped, and a textual feature list is outputted. The more data is put into the tool, the higher the accuracy of the output data.

Web scraping

The last solution of allows you to extract information from websites. In most cases, this would regard websites from suppliers or competitors. You are able to configure the algorithm per website and again, this is convenient for bulk processing of product data. Screenshots of this function is provided below.

Screenshots of (web scraping function). All data is fictive.

With the enriched data that the smart tool does provide you with, companies are able to improve their product filters. Filters can now be more specific on what you and the customers want to see. In this way, customers are provided with more convenience due to the fact that they can find their products more easily. This contributes to the overall customer experience which could in turn increase the conversion rate.

So, would automating the process of enriching data be beneficial for your organization? Does your master data department spend way too much time on enriching product data, while this process can be automated in order to save time and costs? Then take a look at We are happy to offer you a demo to see how many time and costs would be saved if you automate this process.