Completing an incomplete product feed
Posted on March 22, 2021 (Last Updated: March 22, 2021)
We’ve put together a few tips so you can get hold of missing data and complete your product feeds to sell more online.
1. Use info from content databases
One downside of content databases is that they may not have all the information you need. Because these databases were created originally for the logistics industry, they won’t have flashy images or marketing-friendly product titles.
Information commonly found in a content database includes:
- Height, width, and depth
2. Make use of information you already have
Before starting, consider your technical capabilities. Your resources and needs will determine whether this process is more automated or more manual. With WakeupData, you can create rules and filters to automatically and dynamically pull data from different attributes and create a new one.
3. Use automated mapping tools
Mapping Google product categories can be very tough work. All categories must be entered exactly as in Google’s current taxonomy. WakeupData makes this step much easier and more automated.
The easiest way to fill in Google product category information is to use category mapping. For WakeupData users, there are a few simple steps to complete the entire process:
- First, choose where to pull data from. For example, which attribute has information that will most likely be related to the product category? Title and product type are common choices here.
- Map the Google taxonomy that should replace the data from step #1. This is done using a drop-down list of options taken from the official Google taxonomy.
- Then, all you need to do is apply the changes. That means the correlations you have made (linking one attribute to one Google product category) can now run for the entire feed (or even all feeds!).
4. Use a web crawler
WakeupData offer an on-demand web-crawl tool which lets you extract valuable information like unique IDs, tags, images, metadata, etc from across the web - crawling manufacturer sites and databases.