Understanding Customer Search Intentions in Fashion eCommerce
Jess Erdman, February 2021
How would you describe the fabric that the model is wearing below?
To some, the model is wearing a scarf. Or it could be a bandana. Or a kerchief.
Depending on regionalisms and experience with fashion products, you may have one or more words to call a scarf.
While diversity in language has allowed for rich, accurate product descriptions, it has also created a problem for retailers and customers: a confusing, unclear fashion taxonomy.
Fashion taxonomies are associated with text-based search queries and can be limiting for users. For example, if you had searched for the term “scarf” instead of “bandana,” depending on the retailer, you’d be left with empty search results.
Not every customer is an expert in the language of fashion products--and your search system shouldn’t expect customers to be experts.
And, for retailers, fashion taxonomy creates an unmanageable system for classifying product information.
At YesPlz, we have a solution that solves the problem of fashion taxonomy and answers the questions of what users really want: a visual Style Filter that allows customers to demonstrate the attributes they’re looking for on a model.
As a fashion retailer, it’s key to understand fashion taxonomy and how fashion taxonomies affect your customer search behavior. In this guide, we’ll outline different user search intentions, and how you can better meet those needs.
We’ll go over:
In general, taxonomy is the science of naming, describing, and classifying items into categories. Taxonomy is essential in fashion because of the large number of products inherent in the industry, and it points us in the right direction to help us find products.
In theory, fashion taxonomy is an inevitable evil for every eCommerce retailer to conquer--how else would eCommerce retailers even begin to categorize the multiple attributes related to each item?
However, fashion taxonomy has become unmanageable for retailers to keep up with. If fashion taxonomy is defined as a “living, breathing thing,” there’s no way for retailers to keep up.
While taxonomies make sense in static situations (such as the animal kingdom), it doesn’t make sense to constantly catch up with ever-evolving fashion terminology.
And, it’s likely that in trying to play a game of catch-up, retailers are still just a little bit too late to the terminology--after all, no retailer has time to search all of the internet for up and coming terms.
An indecipherable fashion taxonomy is closely related to customer search intentions. When a customer searches for a “bandana” instead of “scarf” and receives mismatched search results, it’s a failure on the part of fashion taxonomy.
When customers choose which words to write in the search bar, there’s a world of intentions behind their choices.
Some customers have a specific problem they’re trying to solve (“dresses that cover shoulders”) or a specific feature(s) in mind (“satin, button-up A-line dress”). Each of these customers has a different goal, and it’s vital that your search system recognizes the differences.
Unless your customer uses the exact terminology that aligns with product tags (more on that below), they may miss out on products--or an entire category.
Every customer fits into a different “search persona”--and some are a combination of different personas.
According to Baymard, there are 4 key search intentions your eCommerce website should be able to support. Below, we’ve created a search persona to accompany the definitions.
1) Exact: The customer has a specific goal in mind
2) Product: The search is intended for a product category, but not necessarily a specific brand.
3) Feature: The customer is looking for a special feature or detail.
4) Thematic: The search is based on a “theme” such as a special occasion or season.
Can your search system support all of the above searches? Hint: Thematic is the most difficult.
This customer knows exactly what she wants, and takes no time to hesitate when typing in her search terms. She can identify the exact brand, dress style, and season. The customer has a specific idea in mind and expects to see an exact match in the result--and might be short on time.
Where Fashion Taxonomies Fall Short: Because of the specificity and speed of the search, the customer may misspell the brand name or mistype a model. Unless the retailer’s fashion taxonomy is mapped in a way to recognize misspellings, she may not receive any results in return. As a customer on a mission, she’ll likely bounce to a different website to continue the search.
This customer has a product category she’s interested in (dress pants, for example), and is looking to explore the category of search results. She wants to see relevant results (and perhaps product suggestions) that complete the search.
Where Fashion Taxonomies Fall Short: Our “willing to explore” customer is likely to become victim to fashion taxonomy terminologies. Take the example above--” dress pants.” There are many different regionalisms and ways to interpret the term, and if she doesn’t know that the retailer calls them “trousers,” she may receive no search results--or, a combination of dresses and pants.
The customer is all about the details--after all, that’s what makes fashion interesting. She wants a cropped vegan leather jacket, under $200. Her expectations are high-- most major fashion brands highlight their clothing features, and she expects nothing less in the eCommerce experience.
Where Fashion Taxonomies Fall Short: As we explained above, fashion taxonomies require constantly updating product features--and an understanding of relevant terminologies. As more features are added to the search, it’s more likely that one of the features that our Fashionista is looking for won’t be named correctly.
The customer is looking for the perfect dress to wear to a wedding--or perhaps a cozy fall jacket. These are examples of thematic searches, and are difficult for retailers to capture, but are an example of how we, as humans, think. We intrinsically know that a casual denim dress isn’t wedding-appropriate attire, but that’s because, just like the themed-searcher, we’ve mapped out those relationships in our minds. The themed-searcher is the most difficult to please--46% of the top eCommerce websites weren’t able to meet thematic search needs.
Where Fashion Taxonomies Fall Short: Fashion taxonomy is, at its core, a system created by humans. Unless the humans behind the taxonomies have considered every possible relationship (which changes based on demographics such as gender, location, economic status) between occasions, seasons, and clothing, it’s impossible to map out all of the possible relationships. Therefore, it’s likely that the thematic searcher will receive few, if any, search results and assume that the fashion eCommerce site doesn’t have the appropriate products for her needs.
Fashion tagging and fashion search go hand-in-hand. When a customer types her search intention into a search bar, she expects to receive accurate search results. But, behind the scenes, inaccurate product tagging can result in a mismatch between search intention and results.
For example, a product tagged as a "fit and flare skirt" might not appear in search results if the customer searches for an "A-line skirt." As we know, the language of fashion can be confusing and inaccurate. When we factor in the fact that third parties are responsible for product tagging, this can lead to inconsistent product tags (after all, we all have different fashion vocabulary). Alternatively, in-house merchandising experts may also make mistakes.
How can we eliminate those mistakes? By using artificial intelligence to tag images, we eliminate human error and increase product tagging accuracy, leading to a rich shopping experience.
YesPlz AI can provide fast, accurate fashion tagging that is rooted in fashion. Our AI is trained to recognize product attributes in all types of pictures, even low-quality images, and is then reinforced with the input of fashion experts.
Here's more about our fashion image tagging process:
Search is at a turning point. Text-based fashion taxonomies are creating barriers for customers to find products easily and accurately.
Understanding your customer’s search intentions and search persona is the first step towards re-imagining search.
However, what’s next?
There’s a solution that surpasses the problem of mismatched search intentions, unclear fashion taxonomy terminologies, and confusing search experiences.
Through a combination of artificial intelligence and fashion visual search, the YesPlz Style Filter allows customers to select the style attributes that best fit their needs (whether they’re looking for a specific product feature, looking to solve a problem, or looking for the winter coat). You can see an example of the Style Filter below:
The future of search is moving away from text-based, static search systems and into AI-powered solutions that take into account customer search intentions.
Fashion AI is heating up in 2022--and these are the top startups we've named that are shaking up the industry, creating new ways to keep customers engaged.
NLP in eCommerce is used to build text-based fashion search for shoppers. But, as any frustrated shopper knows, NLP can only capture part of the language of eCommerce. We explore the problems and solutions to text-based search, and how visual and text-based search can complement each other.
Product recommendations are at the core of the eCommerce search experience. A good product recommendation can make the difference between conversion and shopper satisfaction. In our latest series, we review two retailers' product recommendation systems, and determine which is stronger based on YesPlz criteria.
Your request has been submitted successfully.
Our representative will reach out to you within 24 hours.
Contact us at firstname.lastname@example.org for any further questions.