It would be hard to do the study you suggested in any easy terms, because as the Wichita study showed, only a small percentage (about 6%) chose to use breadcrumbs. So, to create the study, you’d have to create two sites, test in two groups (each site going first), and have enough of a sample size so that of the 6% who choose to use breadcrumbs, you could observe if they worked better. (Back of the envelope math says you’d need to test about 120 participants.)

So, no. Nobody has done that study.

If you’re looking to fund such a study, we’d love to help you spend that money…

It seems that just because people use breadcrumbs (and they don’t very much) it doesn’t actually improve their navigating experience.

http://psychology.wichita.edu/surl/usabilitynews/52/breadcrumb.htm

The question is in fact whether the continuation path offered by breadcrumbs is clear or if there is a more clear way to give them same path?

In most cases, you can do a better job by deliberately designing the paths than by leaving it up to an automatically created breadcrumb system that may or may not be clear to the user.

In either case the SEO outcome is the same. (Actually, by creating better text to describe what that path is, the SEO outcome is slightly better.)

Thanks! You said in 234 words what I was trying to say in 3000 words.

When people get lost they either go back Home to start again (to feel comfortable again), review the options in the primary navigation, hit the back button or leave the site completely to use Google.

I think that every page should have links that users will actually want to follow, and that should be determined by someone thinking about the purpose of the page user is on, and not by automatically cramming all the stuff that is “up the stack”.

If you would think about every and each link in the breadcrumbs, and think “how likely is it that the user will want to go there from this page?” I think most of the breadcrumbs would not pass the test and those links that would pass – probably have a better more prominent place to be.

@Eric: I think the wowhead site has a really interesting approach to breadcrumbs. While they provide location information, they also provide the menu structure in detail. It would be interesting to see what would happen if you applied this to, say, JC Penney’s product hierarchy? Would it still work out?

One site I use a lot has a category breadcrumb system, but they get around this problem by having each of the crumbs being a drop-down menu of siblings, with the current crumb indicated with a tick-mark. This is on a site which extensively rewards mouse-over actions for surfacing linked information so it seems quite natural.

I’ve also seen drop-downs on the crumbs implemented in a slightly different way – each crumb acting as a header for it’s category, with the drop down listing all child-links. They always confused me. The difference between the two models is subtle, but profound.

That said, I think the wowhead site is a lucky exception. Firstly, mouse-overs are extensively rewarded and so users are trained in mouse-hovering, and secondly they didn’t have much of a synonym problem in the first place (it’s a specialist community, and the language terms are already pre-set by the subject publisher).

I would agree that if you’re main navigation and information architecture are doing their job, this type of secondary navigation system is irrelevant for humans.

Happy new year to you and all your readers!

A few factors need to be observed here:

* The time you have available to you as an IA in the business of building (or re-building) a Web site is rarely enough to get the job done perfectly. Your goal is likely to get it (taxonomy/vocabulary/meaning) as close to “right” as possible in the allotted time, especially if a deadline is involved, and often only limited testing is permitted.
* You are involved in a process of Information Evolution. The human mind is unique to each of us and imperfect. Different things have shaped our understanding of the information at hand. That does not mean you don’t try hard to make it all fit for everyone involved. In the end you have what the best in the business have always said is a work in progress. The content and presentation always evolves in a living site. It’s not always perfect or static as we evolve our understanding. It probably never will be. Welcome to the uncertainty principle, language and the human mind. It does get better with time, effort, revision and investment.
* You also have cultural and business-related structural factors to consider. For example: I have little effect over a company using an acronym or cryptic but unique product name to describe the product they’re trying to sell — which despite testing, scores high enough that if you’re new to X product you might not know what it means yet, but you will with a little reading and comprehension. Once you know the product’s name, it sticks with the user and makes a good case for being a part of the navigation of the site in question. So I have to account for that in my IA buildout.

Jared’s definitely right – it’s a cop-out. But I find breadcrumbs are often necessary because any help at all we give to the users of a site to improve “findability” is usually a good thing. Those breadcrumbs can often add helpful context, even at an unconscious level.

Jon’s Kolko’s also correct. Definitely you’ll have entropy, time wasting, emotional drive and the need for connection building to cope with. Our users are often not really goal oriented — or I’d even take it a step further and say that because of the often-scattered human mind, we sometimes lose our purpose for arrival when we scan a site, because other things (like the graphic design or the way the info is laid out or that sneaky well-written, seductive story title in the lower right hand corner) have become more interesting to us. Or the IA was as it sometimes is, imperfect and fraught with human, cultural and structural intervention.

Great article Jared. Thanks!

This might be the cause of your controversy; users commonly visit sites “just to see what it has” in real life, because they aren’t always goal directed in real life. One needs only to observe their own father poking around on a Sunday morning, or visit the library and peek over people’s shoulders, to see this type of casual and non-directionally inclined behavior.

The idea that people are always “task driven” ignores our need for entropy, for connection building, for time wasting, and for the emotional – watching funny videos of animals helps pass the time, but it’s hard to argue that as part of your 94%.