Earlier this week, Peter Morville and Mark Burrell presented a UIE virtual seminar on “Leveraging Search & Discovery Patterns For Great Online Experiences“. It sold out! And I thought Pete Bell and I had done well with our seminar on faceted search!
But I’m hardly surprised. Although I wasn’t able to attend it myself, I gather from Twitter and the blogosphere that it was a great presentation. I enjoyed serving as a reviewer for Peter’s new book on Search Patterns, and I contributed a bit to Endeca’s UI Design Pattern Library while I was there and Mark’s team was developing it.
In reading reactions to the seminar, I was particularly intrigued by a post entitled “Search and Browse” by Livia Labate on her fantastically named blog, “I think, therefore IA“. She raised a question that I think needs to be asked more often: when is (or isn’t) faceted search appropriate?
Her conversation with readers in a comment thread offered some possible answers:
- Faceted search helps users who think in terms of attribute specifications as filtering criteria.
- Faceted search supports search by exclusion, as opposed to by discovery.
- Faceted search requires a set of useful facets that is neither too small nor too large.
I’d like to propose my own answers. Here are the conditions for which I see faceted search being most useful:
- Faceted search supports exploratory use cases, in contrast to known-item search. For known-item search, users are better served by a search box to specify an item by name, or a non-faceted hierarchy to locate it. In contrast, faceted search optimizes for cases where users are either unsure of what they want or of how to specify it.
- Faceted search helps users who need or want to learn about the search space as they execute the search process. Facets educate users about different ways to characterize items in a collection. If users do not need or want this education, they may be frustrated by an interface that makes them do more work.
- The search space is classified using accurate, understandable facets that relate to the users’ information needs. As I’ve discussed before, data quality is often the bottleneck in designing search interfaces. Offering users facets that are either unreliable or unrelated to their needs is worse than providing no facets at all.
Given the above criteria, it’s not surprising that faceted search has been a huge success in online retail: shopping is often an exploratory learning experience, and retailers tend to have good data.
But the success of faceted search in retail overshadows other domains where faceted search may be even more valuable. My favorite example is faceted people search, most recently demonstrated by LinkedIn. I would love to see other entities (locations, businesses, etc.) receive similar treatment, at least in contexts where exploration is a common use case.
I think Livia is right to be skeptical about any interface that introduces complexity–and facets do introduce complexity. I hope that my guidelines help answer her question as to when that complexity is worthwhile and perhaps even necessary to help users satisfy their information needs.
The Noisy Channel 
21 replies on “When Is Faceted Search Appropriate?”
Great to see a post, you’re much quiter now that you’re a Googler.
Faceted Search is appropriate for discovery so I disagree with that one point. Often facets are displayed that people didn’t know about and take users on a different path not just an exclusionary path.
Cheers,
Christopher
LikeLike
Yeah, sorry about the lower posting frequency–blogging has been playing third fiddle to being a Googler and a dad.
I agree with you re: discovery–and I hope that came through in my own second point. But I do think that the commenter on Livia’s post has a representative viewpoint–many people assume that faceted search is about users having fully preconceived needs and progressively narrowing the search space to satisfy them. I think that’s a narrow view of faceted search, so to speak.
LikeLike
Agreed on all counts. 🙂
busy being a dad’s an awesome way to spend your time! 🙂
glad your decreased post frequency is not google censoring you.
have you (can you) say what team your working with or what products?
LikeLike
Thanks for the fine summary, Daniel! And well done contributing to Endeca’s design pattern library. I have high expectations for it.
I can’t agree with you that known-item search is always better served by searching rather than faceted navigation, if that is your opinion.
Sometimes I need to find back to interesting articles I read online. I often end up googling my eyes sore, since I can’t remember the title or any exact phrase from the text, just the gist.
For me, that is a case of known-item search where facets would be really helpful. I know what I want, but not how to specify it (like you said).
LikeLike
Christopher: nope, just busy! Google’s policy is very reasonable: I just can’t disclose confidential information or express opinions that would suggest an official company position. Beyond that, I exercise common sense. My own work is on local search.
Vegard: I agree with you on that distinction and included it in my post about “What is (not) Exploratory Search?“. Sorry if that was unclear here.
LikeLike
Hi Daniel, thanks for posting this!
I am also glad you pointed out “data quality is often the bottleneck in designing search interfaces” because that is another factor I am having to deal with in that context – with bad data even the best discovery pattern falls apart.
Data can be bad in two ways in the context of facets, lack of data (not having the attributes people would be interested in using) or too much data (exposing attributes that people don’t really care about when discovering content). I have both problems to deal with. 🙂
LikeLike
As the only commenter on Livia’s blog who mentioned both exclusion and discovery, I assume you were summarizing my points; so I thought I’d clarify…
I do think faceted search is ideal for exclusionary filtering of search results; i.e. once I have a list of all search results for “brown shoes”, I want to hide everything that isn’t my shoe size. The LinkedIn demo you linked to is a perfect example of this.
As for discovery, Livia was specifically questioning the usefulness of faceted search in the context of video content. In that context, I believe that excerpts (such as a short summary of the video) are more useful than facets for discovery.
On the other hand, I think you’re right that in many instances displaying facets helps educate the user which can promote discovery. E.g., “I didn’t realize there were so many types of chef’s knives”. But in those cases, where facets “take users on a different path not just an exclusionary path”, aren’t we talking about the user abandoning their original search? In other words, it’s search-as-a-shortcut-to-browsing, where the facets are used as improved breadcrumbs.
Search-as-a-shortcut-to-browsing is a great usability pattern, and I think it’s better suited to discovery than browsing-to-filter-search-results (aka faceted search).
LikeLike
Faceted search helps users who need or want to learn about the search space as they execute the search process. Facets educate users about different ways to characterize items in a collection. If users do not need or want this education, they may be frustrated by an interface that makes them do more work.
Don’ t forget the converse! That is:
If users do need and want this education, they may be frustrated by a( non-faceted, sparse, single input box) interface that makes them do more work.
Right? That’s the thing: If you are not in exploratory mode, but the system makes to explore to get what you want, then that’s no good. But if you are in exploratory mode, and the system doesn’t provide you with exploratory tools, then that’s also no good.
I’ve said it a lot and I’ll say it again: I’m glad to finally see acknowledgment of the latter case. For over a decade now in web search, the conventional wisdom is that only the first type of user, the non-exploratory user, exists. And no thought has been given to detect or measure the latter type of user, much less build an interface for that person.
LikeLike
Livia, thanks for stopping by! I’m delighted to have discovered your blog and added it to my feeds.
Dan, thanks for the clarification! In general, I find it very frustrating to search for video content online when I don’t know exactly what I’m looking for. I would like to see faceted search interfaces for video–on everything from Netflix to YouTube) –but availability of useful facets is certainly a bottleneck.
I should clarify that I take a broad view of faceted search. I feel that facets should be used for more than just narrowing queries. For example, a search for [santoku] on a kitchen site should suggest that I consider the broader space of asian knives, which I might then refine by brand, price, type of knife, etc. I realize this isn’t how faceted search is typically implemented. But it should be!
LikeLike
Jeremy, I’ve been one of those frustrated users, so you’re preaching to the converted! And I think that users often go into exploratory mode as they search–at least if they are given the opportunity to do so. I think that the conventional wisdom is evolving, and that now the challenge is at least as much technical as cultural.
LikeLike
Jeremy, I’ve been one of those frustrated users, so you’re preaching to the converted!
What I mean is, I’m seeing a lot of new names in your comments section, and I think it’s fantastic. Awareness is growing, and I like it!
LikeLike
[…] When Is Faceted Search Appropriate? Considerations for using more than just the search box in web interfaces. (tags: search endeca) […]
LikeLike
Hi, Daniel. Could you expand on exclusion versus exploration/discovery?
The use case described by Dan (in the linked post and here in the comments above) is the closest to where I have found faceted search valuable, in product search when I have too many hits for my search and want to toss down constraints to focus my attention.
Is the exploration and discovery limited to suggestions of search terms for exclusion? As in Dan’s example above, “I didn’t realize there were so many types of chef’s knives”?
Or does it enhance exploration and discovery by allowing people to start with a much more general search than they might otherwise (e.g. all AMD processors) and then build a search from the suggestions (e.g. seeing the types of AMD processors, then filtering down to low power, dual core, priced under $150)?
I do see the second use case, but it does require a different starting point where people expect the facets to appear and start with a much more general search than they might otherwise. Is that what you see as using facets for exploration?
Are there other examples of exploration and discovery using faceted search that might be useful to discuss too?
LikeLike
I agree that faceted search offers a great way to narrow results using constraints–it allows you to progressively increase the precision of an initial search.
But I think that faceted search could be just as valuable for increasing recall. Some faceted search applications let you start from a single item and then generalizing to the set of items that share the facet value it has that most matter to you–this is a great way to discover new choices of interest.
A more sophisticated approach would take a set of items (e.g., the results of a search), determine the facet values most typical of that set, and then suggest that set of facet values as a query. Think of it as “did you mean” or “related searches” tuned for faceted search.
LikeLike
Thanks, Daniel, that is interesting. Is what you are suggesting close to what Pandora uses for its recommendations? Finding other items that share an interesting subset of the facets of the original item(s)?
LikeLike
Yes, and not surprisingly I’m a big fan of Pandora. It would be interesting to see an faceted search application on top of the Music Genome data, but I wonder whether users would find facet values like acoustic sonority and major key tonality useful.
LikeLike
On a related note, I often think of the application I mentioned earlier — starting from a very general category and narrowing — not as faceted search but as “browse” (as in browsing a category hierarchy). And I thought of starting from an item or set of items and broadening to similar items more as recommendations than faceted search.
It seems that faceted search might be a bridge between search, recommendations, and browse, capable of starting with a keyword search and narrowing, a high level category classification and narrowing, or a set of items and broadening. Interesting, I hadn’t thought of it like that.
LikeLike
Greg, I wouldn’t just say that it is only “faceted” search that is a bridge between search and recommendation. I think generally “exploratory” search is the bridge. Exploration via facets is but one type of exploratory search.
I tried to taxonomize it in the following link. I’m not positive that my taxonomy is correct, but it hopefully adds to the conversation and gets more ppl thinking:
http://irgupf.com/2010/01/05/more-tensions/
BTW, this is getting right at those debates that you and I used to have back in 2006. Remember, tools vs. personalization? When recommenders meet search, that is where I see “tools” (facets as one example) becoming quite important. See for example:
http://glinden.blogspot.com/2006/04/finding-and-discovering.html
And speaking of the relationship between search, recommendation, and exploration, check out the philosophy behind “Music Explaura”. It’s a good example of that “exploration” philosophy that I was talking about in your blog comments back then:
http://irgupf.com/2009/04/07/music-explaura-exploration-and-discovery-in-action/
In particular, this blogpost is a good place to start:
This link has a video of the Music Explaura in action: (see toward the bottom of the page)
http://blogs.sun.com/searchguy/entry/the_music_explaura
LikeLike
[…] When Is Faceted Search Appropriate? – Noisy Channel […]
LikeLike
[…] When Is Faceted Search Appropriate? – Noisy Channel […]
LikeLike
[…] predefined, high-level categories called facets. While I will send you to Daniel Tunkelang’s The Noisy Channel blog to learn more, I will observe that faceted search is often verticalized, that is, limited or […]
LikeLike