The Future Of Customer Experience Optimization In Fashion: Part 1
One of the solutions offered by CrowdANALYTIX is our autonomous product data onboarding solution that integrates seamlessly with most popular Product Information Management(PIM) systems. Thanks to this solution, brands have sped up the growth of their product catalogs and begun offering their customers much broader product selections, helping retailers compete with even the biggest eCommerce sites that have grown their product catalogs into the tens of millions. Automating the process can increase the speed of product onboarding by 60–70% on average, giving most businesses a quick and significant return.
Beyond this level of automation, CrowdANALYTIX has also helped retailers across industries improve their online shopping experience by optimizing left hand navigation, display taxonomies, product names, product listings and overall product page content to speed up and improve customers’ experience on their sites. One leading American home improvement retailer was able to enrich their product data by more than 93%, resulting in significantly higher conversions.
Like all other retail industries, fashion is relying increasingly on eCommerce, which makes the customer experience paramount. CrowdANALYTIX has developed a new customer experience optimization solution for fashion brands: personalized customer experience optimization, or personalized CEX. This two-part series will explore how personalized CEX can revolutionize online fashion sales by giving customers a drastically improved shopping experience, and giving retailers the ability to sell to and even design to multiple audiences at the same time.
Product Data: Adding Real-Life Detail
The first step in CrowdANALYTIX’ new fashion retail CEX solution is still the optimal structuring of product data: objective product attributes like color, length, size, and many more; the automatic extraction of relevant attributes for cataloging; and most importantly and most difficult, the inclusion of subjective product tags, too.
These subjective tags are what real-life customers actually associate with particular products: the terms that are actively being used in searches to locate and purchase items. For example, while an item’s more objective attributes might include “floor length,” “blue,” “formal,” and “womens,” customers may be searching for it with more subjective terms like “trendy,” “midnight blue,” “clingy,” “fitted,” and “long.”
If a product catalog does not include these subjective attributes, a customer searching for “long trendy fitted navy dress” might never find the product they would be willing to buy, and potential revenue would be lost. A fashion retailer can’t lose anything from the addition of these attributes: the customer searching for “floor length blue formal dress” would still locate the product, too.
Tying Attributes Together
Using AI, CrowdANALYTIX creates what we call “catalog graphs” of products, which track two different aspects of each SKU:
- Relationships between attributes and SKUs
- Attribute values
This ends up being a huge amount of information. It’s an expanded version of what a B2B retailer would need: not just autoclassification and objective attributes, but also product reviews, subjective attributes, search terms, and outfit collections(in other words, which items are presented to go along with the individual product being viewed). Unlike B2B retailers, B2C retailers like those in fashion need to consider subjective elements, because customers certainly do. Individual retail customers often place lots of importance on product reviews or on images of an outfit that their potential purchase could help build.
Note that “search terms” are key to the “real-life detail” mentioned above: by capturing the search terms that bring customers to particular products, fashion retailers can learn a lot about their audience and how they utilize subjective terms to shop. Which brings us to…
Creating Customer Clusters
The idea behind customer clusters is simple, and will be familiar to most marketers: we are simply using AI algorithms to club customers together according to common attributes like past shopping activity, past actual purchases, brands, influencers, and trends. This is just like creating marketing personas, but CrowdANALYTIX is able to use actual customer and shopper data, which includes more amorphous data like how customers perceive and search for items, rather than relying on simple demographic data. Even better, the algorithms we can use to group customers are faster and more accurate.
AI essentially tags customers just as it tags products, and then places customers with the same tags together until it has created 20 or 30 different automatically-generated clusters.
Now that a fashion retailer has detailed structured data in catalog graphs and detailed customer clusters, what can all of this information be used to achieve?
There are three key applications that CrowdANALYTIX can offer that lead to increased customer purchases:
We’ll be covering Search and Marketing in this post, but Design will be the subject of The Future of Customer Experience Optimization in Fashion: Part 2.
Personalized CEX Search
Search optimization is what every retailer wants most. It yields results and improves the customer experience every time it is implemented. CrowdANALYTIX ensures that search optimization includes the personalized information gathered from customers, so that even the subjective terms used by various customer clusters are accounted for.
Not only is search optimized within the fashion retailer’s own site, so that customers get the right kinds of images from searches, and the right product information to make a purchase decision: we also put a focus on external attributes that will drive search traffic to the site from external search engines. This aids conversion from Google etc. just as product information and an optimized site aids conversion from within the brand.
Personalized CEX Marketing
Different customer clusters are likely to respond differently to what is called “romance copy,” the prose descriptions accompanying each item in a product catalog. Although this content might include product parameters like size, color, and category, and might draw on subjective tags from the catalog graph, it is different from these other forms of content in that it can be altered to suit the customer cluster.
Standard romance copy for a pair of jeans might be, “These essential slim jeans are perfect for dressing up or down. With a pale blue wash and distressed detailing, they are a central part of any wardrobe. Pair with a long cardigan sweater for a run to the store, or heels and a crop top for evening drinks.”
This copy can be automatically altered to appeal to, for instance, a woman in her early 20s: “These are the perfect skinny jeans to wear to work or to class. Pair with boots and a sweater for a sophisticated look, or throw on sneakers and run out the door. A neutral blue wash and distressed details go with everything you own for maximum value.”
This might seem complex, but it can be automated completely.
Part 2: Personalized, Crowdsourced Design
Part 2 of this article will cover how CrowdANALYTIX uses personalized customer experience optimization to alter the design of fashion products themselves to increase sales.
This piece first appeared on CrowdANALYTIX.