Comment fonctionne la recherche prédictive et comment l’utiliser ?
La recherche prédictive est une chose que l'on prend pour acquis, mais comment fonctionne-t-elle ? Yext vous montre les avantages pour votre site Internet.
Yext
déc. 13, 2021
Users rely on a lot of tools to help them find what they're looking for — including some they might not even notice. Faceted search systems, artificial intelligence chatbots, and predictive search are among some of the popular ways businesses can help customers get the information they want.
Today, we're talking about predictive search.
What Is Predictive Search?
If a user isn't entirely sure of the data they're looking for, it will be difficult for them to type the correct query into the search bar, making it difficult for the search engine to return what they want.
This is where predictive search helps them adjust their search string to pull up accurate results for what they want.
Online users rely on predictive search functions every day without entirely noticing them. As a result, general search engines have advanced systems that compare a user's query to similar queries and available related content to guide the user to the proper search entry.
Site traffic and predictive search, when appropriately implemented, are powerful tools for increasing those factors. Convenience of use and customer satisfaction can boost conversion rates.
Here's everything you need to know about how predictive search helps your business and your users.
How Does Predictive Search Work?
Predictive search is a function that analyzes popular searches and data within the search engine's index to predict what the user is trying to find and make suggestions to guide them to the appropriate content.
As users, sometimes we don't know the name of what we're looking for or how to find it. Predictive search helps optimize our results. Essentially, it serves as a dropdown of auto suggestions that are generated as the user enters characters into the search bar. Predictive Search, or autocomplete suggestions, are typo-tolerant and can make sense of user intent even if there are misspelled keywords.
The origins of Google's predictive search feature first appeared back in 2004 with Google Suggest, which was then renamed to Google AutoComplete in 2010. This inevitably opened the door to a much more efficient way to encompass an SEO strategy by way of predictive analytics.
In a noninvasive way, predictive search offers the users alternate entry options for what they're trying to type out which they can click on and shift their search to the suggested line.
Streamlining the user experience at the search bar and cutting down on their time spent sifting through search results that aren't what they're looking for is just one of the ways that predictive search benefits users, which in turn benefits your web page.
How Does Predictive Search Help Users?
Predictive search has the potential to benefit users and businesses in a few different ways. Reducing customer search time can get users to what they want faster, which will boost their overall experience on your page.
It can also serve to provide product or data options similar to what they want but didn't necessarily think of initially, expanding the traffic to other content on your site.
Find Users The Best Search String
70% of users report that if a website makes it difficult to find what they are looking for or seems disorganized, they'll leave to a competitor's service and won't look back.
Something online providers want to steer clear of is a zero results page. This is what happens when the user types in a search that doesn't line up with anything on your site, and the search engine doesn't return to them with any data because it doesn't recognize what they're asking for.
Or the user might have a good idea of what they want and get a few of the search keywords correct, but due to the keywords they have incorrect, the search engine brings them content that isn't quite what they wanted, and now they have to look for what they want.
Imagine that you have an online clothing store, and the user is trying to find a specific purple sweater that they saw on an ad, except that it was a purple shirt, not a sweater.
If you have the proper meta tags in place, your search engine will find other purple products, but not immediately the purple shirt they're looking for.
Predictive search is a casual way to provide users with search queries related to content on your pages and thus will connect them that they're most likely looking for.
Predictive search can suggest to them while they're typing to search for a purple shirt instead, which would give them faster access to the product they actually wanted, which increases conversion rates.
Suggest Related Content on Your Page
Most eCommerce websites nowadays have an autocomplete feature in their search option. Predictive search can also be a tool to suggest products similar to what the user wants and give them ideas for other things to check out on your page and enhance the shopping experience.
By using previous searches, the predictive search function can collect and present related search entries as the user is typing. Alternatively, providing a suggested products page later in their browsing experience is helpful, but the consumer may have already finished shopping and is disinterested in anything else.
For instance, imagine you have an online store for car detailing soaps and towels. The user visits your page looking for an interior detailing product and begins to type out their search.
Predictive search might begin to suggest related items such as detailing microfiber towels or different intensities of cleaning products to check out. This is a subtle way of recommending other products on your page and a valuable tactic your business can use to extend product visibility to your users.
Predictive Search Advice
Adding a predictive search function to your internal search engine is a positive step, but there are some extra keys to getting the most of it to remember.
Keyboard And Mouse Functions
Since predictive search jumps into action while your customer is typing, it would be valuable to utilize where their hands already are: on the keyboard.
Rather than forcing them to pause typing and use the mouse to click on suggested searches, make sure that they can use the arrow and enter keys to change their search to one of the recommended options.
Response Time
A very important aspect of predictive search is that it needs to feel like it's running right along with the user as they type. In 2010, Google Instant came on the scene, generating google search results instantly as users type. You don't want the suggested search entries to load up by the time they're done typing, or else they won't be very interested in them.
According to research, the maximum amount of time that will still seem like an instant to a user is 0.1 seconds. When you think about it, this amount of time seems instant anyway, but when loading information, it can make a noticeable difference.
Make Sure That Your Predictive Search is Compatible With Mobile Devices
Over half of all revenue generated by online markets was from users on mobile devices, which means your entire site must be designed for mobile use.
Of course, this includes the predictive search suggestions.
One way would be limiting the amount of extra user interface elements. This would allow the user to focus on their search and the additional suggestions.
You should also make sure that the suggestion entries are at a reasonable text size so that your users can easily read them while typing. Considering readability is important when talking about mobile devices. The font size and line spacing should be appropriate for both mobile and desktop devices.
Furthermore, without the highlight feature, it's very easy for a customer to click the wrong result. For effectiveness, you must support keyboard navigation in your predictive search results.
While there are more benefits to having a predictive search than not having it, user experience is far more valuable than the search suggestions.
In Conclusion
Customers need a well-organized website, or else they might turn away from you and go to a competitor.
Predictive search is a helpful but subtle tool for your users to locate what they want without having to navigate through menus or reach out to customer support. Because believe us: customers prefer to take care of their visit to your web page by themselves if they can. Therefore, you should casually use predictive search to suggest other related products to guide users towards the most successful search query.
References: