Why structured data is essential for your eBay listings

JEZ BROWNING

When searching for your next online purchase on eBay, you may have noticed a new, multi-image, boxed out view. It’s full of useful information like price, merchant, reviews, and special offers. This effectively gives you your own automated comparison site upon your first touch in the online world. Pretty clever stuff and highly useful, even though some of it might be sponsored content.

This is all down to ‘Structured content’, a buzz word, which is doing the rounds alongside others such as ‘big data’ and ‘machine learning’. eBay is making a solid effort to encourage all of their merchants to ‘structure their data’ which is intended to increase our visibility not only on eBay but across the web, too.

Anna, our chief data scientist - ListSmart

To help demystify and benefit from all of this, who better to ask than someone who studies and relies daily on data in its many forms: Anna, our Chief Data Scientist.

Q: What exactly do we mean by structured data?

A: Structured data in its most general sense is, as the name implies, information that has some pre-defined attributes and organization. When data is structured, you know what it is and in what format it should be. In the context of listing products to sell online, it refers to having accurate information in the correct field, in the right format.

Let us unpack that:
Accurate information: information that is true about the listing being advertised.

Correct field: There are many places to display information about a listing on eBay. It is important that the information be paired with the field that corresponds to the type of information it is.

Right format: some types of information have a predefined format, such as numeric or text, or a predefined size, for instance, number of characters in an ISBN number. It is important to use the “standard” way to display the information, in many cases the format will be defined by the manufacturer.

eBay Structured Data - ListSmart

Q: Why is it important?

A: The purpose of structured data is to bring order. By using structured data, both human and machine reading of the data is simplified, because no resources are needed to understand what the data might refer to. Structured data removes uncertainty.

Q: How is structured data used across the web?

A: Structured data in on-line marketplaces serves a dual purpose: it helps both humans looking at the listing to understand the offering better and it allows the site’s search engines to group similar items together.

Usually structured data will come in a “key”/“value” format, where the “key” defines what the data that is contained in the “value” field is and the “value” is the corresponding information. An example would be “key”: price, “value”: £1.

eBay Structured Data - ListSmart

Q: What does eBay view as structured data?

A: Some fields you wouldn’t necessarily think of as structured data but actually are, include : “price” and “shipping options”. Their format is tightly defined, you know exactly what data is in them.

However, the majority of eBay structured data is contained in item specifics and variation fields.

Number of items, EAN, ISBN, model number are all examples of structure data fields.

Q: Why should the seller be pleased it's there?

A: Many online marketplaces use structured data to streamline the buyer journey and help them converge faster on the product they want to purchase. Online marketplaces do this by including filter values that can be used by prospective buyers to narrow down search results. In eBay, the information that defines whether a listing is filtered out is in the listing’s structured data (item specifics, price, shipping options). If a seller has filled out the necessary item specifics, their listing will continue to be shown after filtering by the user, increasing the chance of it being purchased.

Further, if the listing contains specific information, for instance a model number or an EAN, the buyer can be confident that they are acquiring the exact item they desire.

eBay Structured Data - ListSmart

When listing on eBay, some fields can be completed with the information a seller thinks is relevant, for instance, titles. However, eBay also encourages the use of category item specifics, which are key-value pairs. In some categories, the item specific keys are suggested: this serves as a template of what information is relevant in the specific category the item sells in and the sellers can take advantage of this to provide relevant information.

Using structured data, also allows sellers to easily compare their offering with similar listings that have the same value for a specific structured data key.

Q: Speaking of which, what are interesting applications and the future of structured data?

A: With the dawn of big data, we have access to large amounts of information. Some of this data is already very tightly structured: GPS coordinates and timestamps for instance. However large amounts of data are unstructured, for instance tweets and images. Humans are very good at seeing structure in chaos and understanding the implied formats ( 10000 individual pixels make up an image), but machines are not. Or at at least they used to be very bad at it. Artificial intelligence can be used as a way to try to extract the implicit structure into the explicit, making data usable.

Facebook have an artificial intelligence system to automatically caption photos and automatically add some structured data to them. Their platform is very image heavy so they have been working on enabling the Facebook app more usable and inclusive for blind and visually impaired visitors. VoiceOver (a built in screen reader on Apple devices) would normally state basics such as “a picture” or the picture filename. With structured data it will read out “Image may contain: two people, smiling, sunglasses, sky, outdoor, water”.

eBay Structured Data - ListSmart

Google has been working on something similar, you may have seen the results when using image search. Both are just the start of things to come and yet to be fully tested. However, still the most accurate way, nowadays is for humans to enforce structure and take a load off the computers.

ListSmart uses advanced machine learning techniques to automatically group similar eBay products creating our own internal data structure using both structured and unstructured data from eBay. We do this in order to extract insights about best practices and trends and give users the best possible recommendations on how to improve their listings.

Q: What actions can you undertake to make the best advantage of structured data for your eBay listings?

In the meantime, you can sign up to our newsletter to stay up-to-date with the latest news from ListSmart.

Updated: 2nd February 2017

For more hints and tips about selling smarter on eBay, take a look at our other articles and sign up to our newsletter to receive ListSmart’s latest news and articles straight to your inbox.