Designing efficient SEO strategies has always been essential for retailers. Rich, detailed content is one of the main traffic drivers for eCommerces. Therefore, in order to enhance site searches and increase sales, retailers should consider implementing smart product description strategies.
An image is worth a thousand words
In site searches, results are always as good as site data. This means that whenever the site contains poor data, results will be insufficient and they won’t probably meet visitor expectations.
We know that consumers are 80% more willing to engage with content that includes relevant images. Therefore, mapping images to rich attributes is essential in order to display relevant results within a website and show products the customer might really be interested in.
This is even more necessary in the case of eCommerce, considering that product overview and appearance are important factors for a purchase. Actually, 93% of consumers state images are the most relevant factor when deciding whether to buy or not.
The magic of automated, accurate descriptions
When launching a search to find a specific product, we find it annoying to be taken to misleading results. For instance, when we type “red mini skirt”, we do not expect to see pictures of blue skirts, or long-tailed skirts. Neither do we want other items such as red pants to appear among our results. As of now, search engines rank images taking into account several factors. These include the file name, the text surrounding the image, the images alt tags, how the item is categorized, etc. As we’ve seen, the more specific the description is, the greater the chances are to correctly match a search.
In this context, Artificial Intelligence (AI) solutions based on image recognition have started to make their way into the industry. What they do is they generate deeper and more accurate attributes from product pictures in an automated way. Simply said, it is like having a troop of product sheet writing specialists in-house.
But what do these AI solutions consist in?
AI solutions based on image recognition are designed, among other things, to recognize the shape and size of objects. For the fashion industry, this means, for instance, identifying:
- A piece of clothing (i.e. shirt, skirt, trousers…)
- Its colour
- Its fabric (i.e. leather, velvet, denim…)
- Its pattern: (i.e. checked, stripes, floral…)
- Its shape: (i.e turtle-neck, short sleeves…)
The aim is to be able to do so in a similar way a human brain does. Although we are still not able to fully understand how our minds process images and interpret them, certain AI-solutions get relatively close to it and are able to classify and extract very specific information from an image. From a product picture provided by the customer, the machine uses algorithms to recognize specific patterns and draw conclusions. It is also able to assimilate previous experiences (it sort of “learns” by itself).
Homogenising product attributes and categorization
As previously seen, eCommerces are faced with product onboarding and categorization, which are highly time-consuming. When left to humans, these tasks may be associated with typing errors or duplicity in categorization. These mistakes can highly affect SEO ranking, as not only Google takes into account linguistic accuracy, but it also values proper classification. Moreover, the more specific the product attributes are, the easier to match long tail searches and show more relevant products to the users that launch a query. Robust image recognition solutions based on AI help avoid categorization mistakes and provide richer attributes. This streamlines online catalog management and paves the way for search engines to find the appropriate products. And there it is: automated content enhances product discoverability.
Multilingual content for global reach with a local focus
eCommerce sites highly depend on organic search traffic. Optimising images for SEO is one of the actions a retailer can take in order to increase it. However, it can be a critical task when performed in several languages and for huge amounts of images. Machine learning solutions based on image recognition also allow retailers to carry out product onboarding and tagging in several languages. This improves SEO for strategies for retailers in the different target countries and reduces time to market.
Catchoom’s Solutions for Fashion eCommerce
Catchoom’s recently launched solution DeepProducts is designed to enhance catalog management, improve conversion rates and redefine user experience. It uses Artificial Intelligence and Machine Learning technology trained to adapt to the specific requirements of the fashion market.