November 12, 2020
You’re walking into a department store, and you want to buy a pair of nice boots. Naturally, you’re looking for the section with shoes. Once you get to this section, you can easily find the boots you were looking for.
However, what happens if the department store isn’t divided into product sections? What if, for example, they put your boots next to the midi dresses? Would it be easy for you to find them? Probably not. You’d lose a lot of valuable time looking for the boots in the unorganized department store. You’d probably be annoyed because things are so complicated.
Well, the same thing happens with online stores. Even though it seems like everything’s much simpler online, if products are mixed up, it would be challenging for users to find them.
That’s why, when they visit a particular e-commerce store, they expect to see product categories and sections that will make it easy for them to find what they are looking for. If they are looking for a dress, they will look for it in “women’s clothing,” and then “dresses.” It would be odd if you put dresses in “men’s footwear,” for example.
Forrester’s study showed that "poorly architected retail sites" only sell half as much as better organized sites. With a simple category definition and clear product features, there are various approaches to counteract this disorganization.
To make it easier for e-commerce visitors to find what they want, you need product taxonomy. Organized product taxonomy can help them navigate through categories and lead them to the product page they are looking for and checkout easily.
The word taxonomy comes from the Greek word taxis, which means arrangement. Taxonomy stands for arranging, categorizing, grouping, and also recognizing the meaning behind the classified items.
Product taxonomy is a systematic way to categorize, organize, and classify products. That's why it's necessary to understand the products and their context to showcase them coherently.
A hierarchy in which items are classified by use or context (like women's clothing or party dresses) into too narrow subsets is the most common taxonomic model in e-commerce. Presenting the product catalog in segments makes it more manageable and promotes logical connections between subsets.
For example, if the desired product is a red dress, the categories would be:
Home > Women’s clothing > Dresses > Red dresses
You can also have other sub-categories, like evening dresses, casual dresses, silk dresses, short dresses, long-sleeve dresses, etc.
What is essential is that they are logical and based on your understanding of customer behavior. A good retailer should not just put a few categories that sound good together. The best website design in the world will be useless if you don’t structure your website in a way that enables visitors to navigate smoothly. Your product categories should make sense to the customers, making it easier for them to reach the checkout page.
Most e-commerce guides divide website visitors into searchers and browsers. Browsers are generally here to look around. Just like when you enter a store to see what’s in there, without any specific idea of what you want to buy. You’ll buy only if you find something that you like.
Searchers, on the other hand, are here to buy something specific. They know what they are searching for, and they aren’t looking around. They are the ones that demand a well-organized product taxonomy. On a website like this, if they can’t find exactly what they are looking for, you’ll be able to display them items related to their search. Instead of missing a conversion because a particular product is not available, you can convert by offering them something related.
Moreover, taxonomy can have a significant impact on your business decisions. It can provide data and analytics that will highlight bestsellers and eliminate products that aren’t selling enough. Specific categories and proper tagging make it much easier to have comprehensive stats through the entire product catalog.
Findings are more accurate because it's clear what each product is named and what its position is. Consequently, choices that affect which items to feature or remove would become more data-driven and efficient.
Taxonomy is no exception from the digital transformation that affected almost all industries. Data science and artificial intelligence (AI) are technologies that are making it better and smarter. AI in e-commerce can enable retailers to create detailed product descriptions that make product discovery much more comfortable. What is more, it automates and simplifies processes, eliminating manual actions and allowing retailers to focus on critical business decisions.
Competitive companies have started to leverage these advantages. Some of them use the power of AI to identify user profiles and the products they are interested in, displaying them recommendations based on their preferences.
Others have developed AI-based engines that automatically generate product descriptions and content. For example, Facebook’s GrokNet has the ability to detect and describe the products in pictures and therefore help sellers sell them on the Facebook marketplace.
Another challenge that AI solves are linguistic differences. Different people use different words and terms to describe the same product. Online retailers need to be careful and consider all possibilities, which is something that AI-based systems can address to provide a seamless shopping experience.
Online stores with catalogs that have thousands of products need more complex and scalable solutions. They need intelligent systems that can replace manual work hours and process catalogs in a few minutes, without being prone to errors.
Automatic tagging is an AI-based solution that removes the need for manual product tagging. In the product catalog, automatic tagging organizes and marks images based on their attributes, using sophisticated AI algorithms. These algorithms speed up the tagging process and eliminate the need for human interaction, thanks to Deep Learning.
This is essentially a method that creates catalog asset metadata. It operates in a way that examines the image and identifies characteristics linked to specific keywords.
The automatic tagging system’s foundation is trained AI algorithms that can recognize clothes in pictures just like we humans can do. We only need a millisecond to determine whether we’re looking at a blouse or leather shoes. For machines, this is a bit more difficult, and that’s why people have so far performed tagging manually.
However, now that we have Computer Vision and Deep Learning, things are different. These technologies enabled scientists to create Neural Networks that imitate the actions of the human brain. Trained Neural Networks can identify objects in an image just like humans can do. They take a picture, scan it, and provide semantic information about the items presented.
To make it easier for algorithms to identify objects, they need to process many images. This way, they learn how clothing items look like and what terms can be used to describe them. Basically, this is what automatic tagging presents. An AI-based system that has processed thousands of tagged fashion images to become trained and to gain the ability to tag fashion images with characteristics and categories independently.
The Deep Learning algorithm processes visuals’ content, such as pictures or videos, captures their characteristics and discovers objects. An efficient automated product tagging model can reduce the processing time of catalogs by up to 90%.
Every image gets many tags that offer in-depth and detailed product knowledge. When the system gets trained, the tags start collecting information about how they are being used, who is using them, and their connections to other tags.
As products get a number of tags, a skirt can have the following tags—”black skirt,” “A-line,” “midi skirt,” “leather skirt,” “formal,” etc. This means that if a shopper is only looking for a leather skirt, the AI engine will consider all other attributes. If another shopper is looking for a black skirt, they’ll both be able to find the same skirt in the list of products.
Aside from the fact that it eliminates manual tagging, automatic tagging provides more accurate product tags and contributes to the increased quality of search engine results. Moreover, it can significantly reduce operational costs and time-to-market of new products, making the entire digitization process automated.
When you enter a nicely-looking physical store, you feel good. Everything is organized and visible. You can find what you want quickly. This is a good user experience.
The same thing happens with an e-commerce store. When the customer lands on the website, they want to see a nicely-looking, clean, and organized interface. When they are looking for a specific product, they navigate a broader category into a more specific category containing the product page.
Obviously, a categorization that might sound rational to you doesn't have to be the most logical one to the consumer. So, before you create your product taxonomy, it is desirable to put yourself in your target customer's shoes. Test different paths and hierarchies and document the necessary steps that lead to the product page. You should strive to minimize the number of steps and make sure your product taxonomy is as simple as possible.
The key lies in balancing the rational associations between your products and the consumers' inherent impulses, which are mostly, though not always, compatible. Keep things simple. This should keep them longer in your e-commerce store. At the same time, simplicity and traditional categorizations should create a familiar experience and avoid confusion.
As website visitors can be browsers or searchers, they either know exactly what they want, or they are only searching and comparing choices. Therefore, they would either investigate the categories of the e-commerce website or use the search bar. This means that product classification and quality search results are essential to the customer’s happiness and, consequently, conversions.
As customers don’t always have the same perception of where the product should lay, they should have the option to find the product in both the most narrow child category and the broadest parent category. For example, if a man is searching for leather boots, he will go through Men’s footwear > Boots > Leather boots. However, he should also be able to see the entire offering within the parent Men’s footwear category.
Next, the overall navigation experience of the website should be intuitive. For example, on the mobile view, the hamburger menu should lead to the entire product categorization. What is more, you might want to add your most popular product categories as featured on your website.
For a more personalized experience, once the user logs in to their profile, they should see the product categories customized to their preferences. These could be the categories they already purchased from or others that are related to previous orders.
Although, as the owner, you should be the one who knows the products best, you should also be able to share all this knowledge with the most critical stakeholders. Your team members and your customers are the ones who need product information the most.
That’s why you should focus on the crucial information, starting from the top products of the product catalog. Extract the most important information from these products and build the product taxonomy based on it.
To perform this more effectively, you can assign product owners. These will be the team members responsible for specific products. They should identify the essential characteristics of a product, what makes it different from other products, and use it to make taxonomy smarter and more intuitive.
This should help you identify consistent product transparency and close the difference between how the sales staff can position the product and how the product owner does it. This can also help you decide which item corresponds to a category and what features you want to link to a particular product.
Each search engine results page (SERP) requires facets that help you filter down your results by separating characteristics. Clearly, distinguished product attributes return more quality search results and provide customers with the exact products they are looking for. If someone only wants to see red boots, why show them black boots as well?
Thanks to product taxonomy, your search engine filters will be more precise, and the search results will display the correct products.
Taxonomy can also help you to rank better on Google. With clean and structured data, Google’s crawlers will quickly identify your content and rank you according to the accurate keywords.
To be able to bring smart business decisions that will have a positive business impact, in the long run, you need to have deep insights into how your products perform. You should monitor both your bestsellers and your worst performers.
Product categories will enable you to easily identify the products that don’t sell and eliminate them. On the other hand, it will highlight the bestsellers that you should feature on your home page. This will lead to better, data-driven decisions in real-time.
Although product taxonomies are different for each store, here are some tips on how to prepare for the process:
The team should be familiar with all activities regarding product taxonomy. After all, this will change the way you work in many ways. This way, they will be more open to changes and prone to adapt quickly.
With good categorization, things will be easier for both employees and customers. If users can easily find their desired products, they’ll have a happier and more memorable user experience. This means that they’ll come back again and become loyal, benefiting all company employees.
Elaborate on why product taxonomy is crucial for your e-commerce store. Your employees should understand that your customers might not have the same perspective of your products as they do. Show them examples that will make the need for taxonomy clear and be transparent. This will be a significant process, and your employees might need more time to adapt to it.
Taxonomy is all about bringing your products closer to the users. This means that before starting the process, you should get to know your users profoundly. Psychology plays a significant role here. Learning how users behave and what terms they use is crucial to performing product taxonomy well.
Test various versions and monitor how they navigate through your pages. Collect data on what they search to study the language they use. Consider different approaches to how your products can be perceived.
Sometimes it’s also better to do some informal talking to people who suit your target group. These conversations might uncover things that numbers can’t. Let them describe how they feel about your website navigation and ask them to show you how their favorite online stores look like.
Remember, the customer is always right. You can’t give them a complex store just because you like it that way. They should be able to find products easily, on their terms.
Make sure you know how much content you have on each product. Have it easily accessible and understand your content sources. When you have a better overview of all your information, it will be easier for you to organize it. You won’t encounter a situation where you’ll be missing data for a particular product, and things will run more smoothly. Focus on the most critical content that helps you make better decisions.
Product taxonomy has to be user-friendly and straightforward. If you create too many categories, people won’t be able to find the products they want to buy. Complex taxonomy simply doesn’t work. When coming up with the categories, don’t go too deep. It’s better to keep more broad and shallow categories.
Hierarchies shouldn’t be split into more than two or three categories. A user wants to see things clean and clear. But, they don’t want to spend hours navigating through your endless number of categories. Not only does over-categorization confuse the buyer, but it also increases the risk of an item being classified in more than one category.
Be careful with product names, and always keep your user in mind. Be careful with names and, at any point, remember the customer. To encourage search, use words they can comprehend and utilize synonyms. Try not to use jargon extensively.
Also, try to fit all products into a particular category, avoiding categorizing them as “Other.” Users rarely search for products here.
Hierarchy makes things more clean and organized. For example, in written articles, we have titles and subtitles to mark the paragraphs’ hierarchy in the text, providing structure for the reader. In e-commerce, taxonomy is the one that provides the structure for the buyer, dividing the products into logical categories.
Great product taxonomy makes it easier for search engines to locate products and consumers to browse, identify, and add products to the cart, effectively leading to conversions.
The taxonomy should enable users to find what they want faster and with less effort. Technologies like AI and data science make the entire taxonomy process automated and smarter, saving a lot of time and online retailers’ work.
Pixyle.ai’s automatic tagging engine can automate your entire product taxonomy process. Get smart tags that clean up your catalog, improve your search engine ranking, and boost your conversions. Scan your entire product catalog in a few minutes and eliminate manual tagging that can take for weeks.
Harness the power of Visual AI