In Part 1 and Part 2 of this series, we gave some tips about optimizing your search box and outlined some key features of the high-performing autocomplete that every e-commerce store should consider. The next step in the funnel is taking your customers to search results page and giving them what they’re looking for! They’ve taken the first steps towards finding products they are interested in, now your search needs to deliver.
1. Deliver super relevant results
Your search results page should, first and foremost, provide only relevant and accurate search results. Sifting through irrelevant products is tedious and confusing for your customers. Search typically works by way of matching the text of the search query to the text in your product descriptions; if the search finds a match, it displays the product in the search results. But text matching is not always the best indication of relevancy, so other indicators of relevance should come into play to take your search to the next level.
By feeding your search algorithm with huge amounts of data generated from your customers in your store, your search can continuously learn, adapt and improve. Search solutions that use advanced machine learning will offer you this level of intelligence.
2. Handling multi-word search queries
With most search solutions, single-word search queries will often give relevant results without a problem (it’s the least they could do). But multi-word search queries are more complex and your search needs to be a bit more intelligent to handle them. If a customer searches for “blue suede shoes for men” they don’t expect to see all blue products, all suede products, all men’s products and all shoes. A smart search will show results for men’s shoes that are blue and made of suede.
3. Understanding synonymous words
People use different terms and expressions for different things; I might say trousers you might say pants. Your search should understand this, then learn and deliver results accordingly. If all of the products in your dataset are labelled as “earphones” but your customers are searching for “earbuds” or “headphones”, your customers will get relevant results no matter what your customer searches for, instead of reaching a dead-end 0-results page.
4. Understanding different forms of the same word
Having a “stemming” feature in your search functionality allows for a more natural search behavior. “Stemming” allows your customers to search for terms like “watch” or “watches” and get the same results. Or “running”, “runner” or “runs”, identifying them as having the root of “run” and returning all results with this understanding. It can be as simple as giving results for a “shirt” based on the given plural query of “shirts”.
5. Handling typos seamlessly
Originally Google’s brainchild, “search instead for” and “did you mean” features help your customers avoid the 0-results page whenever possible. Your search should recognize the typo and automatically provide results for products matching the correct spelling, displaying text like “0 results for ipohne, showing 20 results for iphone instead”.
6. Offer filters for fine-tuning the search results
Now that you’ve returned super relevant and accurate results to your customer, they may want to narrow them down further according to more specific criteria. These filters are a tool for the customer to fine-tune their search, getting them closer and closer to exactly what they have in mind. It’s equally as important to make sure your filtering options are relevant to the products in the results, i.e. a filter for “shoe size” should never appear in results for a “dress”.
7. Offer partial matching
It is very common that shoppers over-constrain their search results by including too many variables in their query, which will in many cases lead to the dreaded 0-results page. Partial matching is when your search gives results for one or more but not all of the terms in a multi-word query. Clearly indicate for which terms there are results and which terms have been omitted. Partial matches is also helpful because shoppers will adjust their query with the clues you’ve given them on how to reach their target results.
8. Optimize your 0-results page
“Search, more than any other activity, is a living, evolving process of discovery — a conversation between a customer and the website. Unfortunately this conversation is often fraught with miscommunication, and so it is critical for you to keep this conversation going even when the customer has initiated a search that yielded no results”. In Greg Nudelman’s book Designing Search, he gives a concise 0-results strategy with these 3 design principles:
a) Clearly indicate there are no search results, so your customer can recover. Don’t be afraid to say “I didn’t understand”, don’t try to hide the 0-results condition, that will only confuse the shopper.
b) Focus on providing your customers with a way out. Your 0-results page should never be a dead-end that makes the customer feel like they have made some sort of mistake. Every control on the 0-results page should do something to help get the customer out of the 0-results situation.
c) Provide multiple strategies within your 0-results page to recover, with links related to the spirit of the original query. If the query is over-constrained by too many variables, partial matching might help (see “#7 Offer partial matching” above). Other common methods are to display or link to “featured products”, “most-popular products”, relevant third-party resources (like contractors), or relevant advertisements.
This list is not exhaustive, but depending on the quality of your current search it may be a lot to consider fixing all at once. But taking care of any one of these points will already be a great improvement to you search quality and the overall experience for your customers. It’s always important to remember that customers who use your search are more likely to make a purchase, so optimizing it will definitely pay off.
This post is the last post in a three-part series about optimizing search for your e-commerce store, read the introduction “A Three Part Series: The Essential Guide to Ecommerce Search”.
In “Part One: 6 Tips for Creating the Best Search Box”, we give some advice on optimizing your search box to attract your customers to it. This is important as we know that customers who use search, are more likely to make a purchase.
In “Part Two: 6 Ways to Optimize Your Autocomplete Suggestions”, we outline the essential design patterns and features of your autocomplete overlay. A common search tool, autocomplete suggestions are nothing new, but you’d be surprised by how the small details are the difference between a good user experience and utter confusion.