When you think about web design, on-site search is probably something you see as a simple bar with a magnifying glass icon. Actually, this is just the tip of the iceberg. Think of all the times the search results in online stores disappointed you. On-site search is a feature that requires thoughtful design and high technical quality which, if not taken care of properly, can cost you customers and revenue.
In an ideal world, every e-commerce operation would closely monitor the number of visited pages and clicks from customers. When you think of such monitoring, you probably associate them with your store navigation and customer journey from home page to checkout: category pages, product pages, cart page, etc. What if you treated the on-site search bar as a separate site, and tracked all clicks and behaviors just like from any other site location? This is the approach you’re going to take when optimizing the search engine in your store. Why?
Because just like with any other mistake at other points in the customer journey, a faulty search tool will cost you money. The search is a component of a transaction cost. That said, it’s a cost worth taking. Customers who use on-site search know what they want (or at least give you a decent idea of it). Responding to their needs with the right communication will be much easier than adjusting the website content to someone who has no clue what they’re looking for.
“Performing an on-site search is a strong behavioral predictor of intent to convert.”
Econsultancy and ConversionXL
But how do you spot elements that will improve your store search? What obstacles and difficulties do you need to prepare yourself for? Here’s a quick guide to common mistakes in e-commerce search engines and the best ways to fix them.
Poor search interface
A technologically advanced search engine will be of no use if you forget to put it in front of your customers and make it work like a no-brainer. Let’s list a few key factors in a well-designed on-site search:
The search box should be easy to spot and big enough to enter typical queries. In the case of search filters, the user must be able to understand each category and tag their desired filters without a hassle. Think of such factors as icons, color samples, category order, etc.
This feature provides several advantages. First, it makes the search process faster and encourages the user to choose the most adequate keyword (which, in turn, will display the best search results). Second, it limits the possibility of typing a misspelled keyword or other incorrect phrasing that needs more processing than the proper phrase.
This may sound like a more complicated procedure but using a multifunctional on-site search tool will make things much easier. What does the on-site image search look like? It lets users upload a picture or insert a URL of a picture online. Based on pixel similarity, the search results should present visually similar products or even a similar one if available.
Such a way to search products is especially handy for users when they’re looking for an item with different names in different parts of the world, like catsuit, overall(s), boilersuit, etc.
Irrelevant product results
It might seem perfectly simple – type in “iPhone X” and see all variants of iPhone X. Well, that’s not the case anymore. Customers expect your on-site product search will answer less precise queries, such as “iPhone big screen” or “smartphone no buttons”. You can thank Google for that, as their algorithms are one of the closest to customer intentions. The most common issues include:
- Misspellings. Go back to primary school for a second and remind yourself of all those spelling contests. Not everyone was a winner, were they? Remember that they’re all using the internet and browsing for products online now.
- Synonyms. Your product descriptions aren’t going to be enough. Chances are there’s another word for what you are selling and someone out there is going to search for it.
- Acronyms, abbreviations, initialisms. The urge to communicate in short forms is clearly visible in online slang and informal language. Make sure your on-site search “understands” what customers may have in mind, especially your younger clientele.
- Plurals, apostrophes and other punctuation marks. These are integral elements of text communication and can’t be left out of your on-site search engine.
Slow loading results
Yes, everyone talks about how important page load speed is, but apparently there’s still a need for emphasizing this issue. It is especially acute on mobile sites, which tend to load slower than their desktop counterparts. It’s even more important after a statement from Google regarding their algorithm updates:
“[...] Starting in July 2018, page speed will be a ranking factor for mobile searches.” (Zhiheng Wang, Doantam Phan, Google)
To minimize the website loading time, you will need to take a close look into several development issues, including the number of HTTP requests on each site, server response time or file size.
Lack of personalization
If you think that displaying results adequate to your customer’s query is enough, think again. There are many other factors that may affect their decision to continue browsing your offer or not, such as product price range, brand preferences, color selection, etc. To provide different search results for different customer profiles, you can implement an AI-driven search engine.
How does it work? Machine learning algorithms use real-time data to deliver the most personalized experience possible. They also use behavioral patterns, historical data for individual customers and transactional data. Every page visit or website click teaches the algorithms to deliver better results. Also, properly developed search algorithms can be compatible with filters and facets in your search interface.
Neglected “No results” page
When users search, some will inevitably end up at a “no results” page for various reasons. But that doesn’t have to be the end of their shopping path. Actually, “no results” may mean “other personalized items you may like”! Everything depends on the model of your page:
- Generic – They display a pre-set selection of items and categories, not taking your customer’s query into account. It’s a good idea to have such a page prepared in case you have practically no data about this particular person. But that’s the only option when a generic page works better than nothing.
- Intelligent – Such pages display product suggestions based on the customer data available. The results may include alternative search queries, synonyms and other ways to improve search results.
As an online store, you work with loads of customer data for different purposes: the delivery process, general communication, marketing and more. Why not put the same data to use to yet another purpose and increase the chances of closing the deal?
In today’s e-commerce, finding more ways to benefit from customer data can help you stay ahead of the competition. A great example this is personalized on-site search. Upgrade it from just a simple search tab to one of the first steps of your customer journey. The more personalized elements you have on your page, the more insights you can extract and keep optimizing.