Search Assistant

The Search assistant is a web component used to enable an intelligent search box. For the search to be available on all pages on a site, it is recommended to add it to the header.

Key features to help visitors have a frictionless search experience include phrase and product suggestions. There is also support for content links and other features.

Search assistant

Phrase suggestions

Phrase suggestions

Phrase suggestions provide the visitor with relevant phrases to the start of their search query, that most likely match their intent. Some of the featured techniques to make optimal use of the screen real-estate are described below.

Phrase diversification

The phrase suggestions provide diverse options that helps narrow down the search, given the user provided input. This means that the phrases are not only well considered suggestions on their own, but they provide clear separate options when used together.

For example, provided the input tsh the following suggestions: tshirt, t shirt, t-shirt, tshirts, t shirts, and t-shirts, might all be good suggestions separately but not when presenting them simultaneously. It is not only unhelpful as well as a waste of space, but also confusing since they all ought to lead to the same product result. By ensuring that the options are diverse, additional helpful information can be added to the different suggestions instead.

The use of infix search allows eSales Fashion to find suitable suggestions even if the suggestion does not start with the search phrase. For example, a search for dress could suggest the category Maxi dresses.

Autocorrected phrases

Visitors frequently misspell words. eSales Fashion will automatically correct to what it thinks the visitor probably means while typing, and at the same time provide phrase suggestions based on the corrected phrase, so as not to penalise the visitor for their typo.

Suggestion stability

Stability of phrase suggestions provide a better user experience. Search phrase suggestions should not necessarily change just because another letter have been added to a search phrase. The phrase suggestions should only change when eSales Fashion interprets the input to change the intent of the search phrase. For example the search phrase tshi should provide the same phrase suggestions as tshir, as the intent behind both phrases likely is interpreted as t-shirt.

Semantic intelligence

eSales Fashion has a degree of semantic intelligence due to its understanding different concepts in the data model, combined with its intelligent query analysis. If a visitor would enter green t-shirt as a search phrase, eSales Fashion will interpret this a t-shirt with the colour green prior to a t-shirt with the title Green Day.

The semantic intelligence also understand that some concepts are good to combine and others are not. A visitor entering skater dress as a search phrase is for example not helped by appending the word dresses. By understanding the product hierarchy, information redundancy can be avoided. Furthermore, eSales Fashion understands what parts of the data likely are nouns and adjectives and uses that linguistic knowledge when proposing phrases.

Phrase highlighting

To enable a visitor to clearly see how the suggestions differ from the entered phrase, a stable visual highlighting is implemented. The highlighting is performed on the part of the phrase that differs from the user input, as this has proved to be more visually helpful in UX research studies by the Baymard Institute.

Highlighting remains stable even when providing corrected phrase suggestions, in these cases the difference of the anticipated meaning of the phrase is highlighted.

Product suggestions

The product suggestions are built to provide visitors with relevant products that, based on their current search phrase, most likely match their intent. Stability and consistency in the product suggestions is highly important for a good user experience.

Product suggestions

Suggestion stability

Stability of product suggestions provide a better user experience. Product suggestions should not necessarily change just because another letter have been added to a search phrase. The product suggestions should only change when eSales Fashion interprets the input to change the intent of the search phrase. For example the search phrase tshi should provide the same product suggestions as tshir, as eSales Fashion interpret the intent of the search phrase to find products based on the phrase t-shirt.

Search consistency

eSales Fashion provide the visitor consistent results for their search term, both in the Search Assistant and Search Results Page. For example, sho may present three shoes in the Product Suggestions, and the visitor expects to see the same three shoes at the top of the Search Results List.

Autocorrected suggestions

Visitors frequently misspell words. eSales Fashion will automatically correct to what it thinks the visitor probably means while typing, and at the same time provide product suggestions based on the corrected phrase, so as not to penalise the visitor for their typo.

A visitor searching for a Product Code will be automatically transferred to the corresponding Product Page, without any additional steps.

Content links

Many visitors are looking for information regarding returns, shipping and delivery, physical stores etc. Allowing them to find this type of information via search is supported in eSales Fashion though the use of content links.

Clutter avoidance

Whilst phrase and product suggestions are presented after a single character has been entered, any matching content links will be presented after three entered characters. This helps the visitor, as it is more likely they are looking for product than content.

Result separation

There is a clear visual difference between what is a suggestion of a search phrase and what is a suggestion for a content link page. This allows a visitor to have a correct anticipation of what will happen when interacting with the suggestion. For example, if a site sells shoes with the brand Contact it will clearly be separated from a Contact page served as a content link.

If an executed search yields no product results, but there is a single content link result matching the search phrase, a visitor will automatically be redirected to that specific page instead of an empty product result page.

Recent searches

Visitors may want retrace their steps when re-visiting the site. The most recent search queries are displayed, so that visitors don't have to re-enter previously executed search phrases.

Clear recent searches

The recent searches can be cleared with a click. An easy way to keep shopping for presents on a shared device a secret, and most importantly a necessary feature when saving the recent searches of a visitor.

Other features

The eSales Fashion search assistant comes with some additional helpful features, designed to assist the visitors and make their search experience as frictionless as possible. Each with a distinct purpose.

View all

UX research has shown that a small percentage of online shoppers may interpret the product suggestions as the entire search result. Therefore it is important to visually complement the suggested products with an indication that more products are found.

Clear search field

Clear the search input field with a click.

Phrase persistence

Visitors often want to adjust their search phrase. On the Search Results Page, the entered search phrase is kept within the search box, so visitors don't have to re-type their original input.

Keyboard navigation

All content within the eSales Fashion search assistant is accessible using the keyboard. This includes moving between phrase and product suggestions with the arrow keys and using the Enter key for executing searches.

×
Copyright

This online publication is intellectual property of Apptus Technologies. Its contents can be duplicated in part or whole, provided that a copyright label is visibly located on each copy and the copy is used in conjunction with the product described within this document.

All information found in these documents has been compiled with utmost attention to detail. However, this does not guarantee complete accuracy. Neither Apptus Technologies nor the authors shall be held liable for possible errors or the consequences thereof.

Software and hardware descriptions cited in these documents might be registered trademarks. All trade names are subject to copyright restrictions and may be registered trademarks. Apptus Technologies essentially adheres to the manufacturer’s spelling. Names of products and trademarks appearing in this document, with or without specific notation, are likewise subject to trademark and trade protection laws and may thus fall under copyright restrictions.

CLOSE