In a previous post, we looked at what entails product information beyond it just being information about a product. A combination of offers and catalog data makes for a comprehensive product listing. However, is that enough in today’s age of digital commerce? Not really. Having all the information available is one thing. But how can it truly benefit a brand or retailer? How can it help a consumer find what they are looking for? To be useful for both sellers and consumers, product data needs to be structured.
The short answer is that search engines cannot read free text. So if businesses want their products to be discovered and bought, they need to conform to a certain structure. In the interest of not getting too technical, generally, product data needs to exist in a attribute=value pair. For example, for a blue shirt, the color would be represented as “color=blue” on an ecommerce website. The format would change depending on what markup language was being used. Let’s look at some examples to make this clearer.
Google for example, has strict specifications when it comes to feeds submitted to Google shopping. But even for products to be discovered on the search engine, the more structured the information, the better it is for both ranking and discoverability. Product title, type of product (categorization), availability, unique product identifiers (GTIN, SKU, MPN), list price, sale price, condition of item, features, attributes, specifications etc. are some of the parameters that help in product search and discovery. For apparel, this requirement gets even more granular as we can see in this example that Google provides.
So, what does this mean for retailers? I looked up a random product on a website I often shop from, www.iherb.com. I took the URL for the Giovanni Anti-Frizz Serum and popped it into Google’s Structured Data Testing Tool (it lets you determine how structured and by consequence, how search-friendly your content is). The product did pretty well except for a couple of errors.
Then I searched for the product within Google (incognito) to see where it ranked and it did show up in both the organic and paid search results. Whether iherb.com offers the lowest price or not is secondary. The point is there was enough relevant data for the product to show up in a search result with just a few keywords.
Keep in mind that retailers usually get product data from brands and manufacturers. So it’s important for the first entity in the chain to ensure that data is structured.
While the example we’ve seen might be specific to a search engine, there’s no denying the need and urgency for structured data. Imposing structure helps to get the right kind of data into the system. Another important takeaway from here is that what we can see lacking here is the presence of a universal product language. We’ve looked at one kind of structure here, but what we need is truly universal structure that is followed by anyone dealing in product information. It will enable pervasive and consistent shopping experiences. That however, is a topic for another time.