Search suggestions matching, what is the best way to bold matches?

Question

I'm trying to understand the pros and cons of different methods to hit highlighting matching text within the search box drop down suggestions. I'm aware of the article by NN group https://www.nngroup.com/articles/site-search-suggestions/ in which they specified good practices such as:

• when matching word is upfront, highlight the completing text
• when matching word is within, highlight the matching word

This is quite clean when we only look at query formulation and query completion. I'm looking at suggestions that support matches on other metadata such as author or storage location to name a couple of them. When the data gets complex I feel the model NN group suggest doesn't help to quickly scan and read through the suggestions.

At the moment I'm looking at 3 potential models

Does anyone have experience with a similar challenge? Which model is the most desired and why? Eventually, are there any other models I'm missing out?

Answer 1

I think that the quoted article does not mean that in one result or sugestion set there should be different highlighting within the same set.

In your case the search is a "like" search (search string is "%cat%"). This means that, you find results where the searchterm can be in the middle of a text. Therefore rule 2 applies highlight the searchterm in the result even if sometimes the searchterm is in the beginning of the Title. this would be your third example.

Highlighting the completing text should only be used if the search is implemented that the result has to beginn with the searchstring, as if you would search "cat%".