Does every site need search?

We're working with a UI/Design expert for Path 101 and we've been having an interesting little debate about putting a search box on the front page.

Here are some reasons you'd typically want a search box on the front page:

- People understand what to do with a search box.  You see a box, you type something in it, you get back what you asked for. 

- Cuts down clicks--gets people right to what they want.

- It has a high chance of engaging users.  Like moths to a light, users will type something in a search box if you put one on the page.

- It gives you data on what people are looking for.

- It is neat and uncluttered, versus trying to put up links to every possible thing the user could have asked for.

However, there are some things you need to think about before following conventional wisdom.

First, what are your chances of getting the user what they want as a result of that search?

Search relevance isn't easy.  You have a lot working against you.  First, no matter what you ask for in that search box, people will undoubtedly type in all sorts of random, irrelevant crap--i.e. stuff you don't have results for on your site.  Search is chaos. 

If your site is about selling a specific product, like a 3/4'' inch bolt, no other size bolt will do.  However, if you're searcher is just looking for "stuff that holds something in place" and wouldn't mind browsing bolts, clamps, glue and other various fasteners, your search is going to have to be pretty intelligent to understand the relationships between those items.  No open source free text search is going to figure that out.  Search promises to answer your question, but don't underestimate how complicated (or simple) the question may be.

So, if you can't provide relevant search results to your users, what's the point of having it?  Let's say 50% of the time you find something relevant for a user.  What do you think happens the other 50% of the time?  I'll bet you that you lose most of those people, because they're assuming that search is pretty comprehensive and that a search that comes up empty means you can't help them.  However, if you had something closer to a site map or directory on the front page, a la Craigslist, not finding something relevant there, in all likelihood, does mean that you just don't have what they're looking for.  However, at least then you have a chance at guiding that user to something else--a higher chance than you would after you get a "Your search turned up 0 results" message.

Does search=a quickie?

What do you want users doing on your site?  Coming in and taking out just what they want with surgical precision or do you want them to sit and stay a while?  What kind of behavior does search encourage?  Sure, you don't want to make it difficult to get people want they want, but I also think you want more than just a millisecond to show someone what else they can do on the site. 

The well informed click

The power of search is that it gives you information about what the user is looking for.  What it doesn't give you is any idea about who the user is--the context for the search.  That would give you significantly more relevant results.  Consider this example...  You go to a healthcare site and the first thing it asks you is if you are a doctor or a patient.  One click later, you have an enormous about of context and information about what kinds of results you should be showing your users.  Imagine if you had typed in "flu" in just a plain old search box, because you had it, but

all the results you got where "What to prescribe your patients when they have the flu?" and "What to charge for a flu shot?"  You'd think the site was just for doctors and probably move on.  I've been thinking a lot about the "well informed click"--the idea that you can setup your UI in such a way that each click tells you something about your user and what they're looking for. 

The real searchers go to Google: SEO as a replacement for your own search box

Particularly at the beginning of your site's existence, most people looking for something specific on your site won't start out at your home page--they'll start out at Google.  Even when I know that IMDB is the most likely destination when I type in the name of a movie, I still don't go to IMDB first, I start at Google.  So that begs the question...  who winds up at your front door.  My theory is that it's two types of users--people who came in via a recommendation and people who typed something general about what your site does. 

People who came in on a recommendation often come through a well informed click. visitors will come from a link in TechCrunch telling them that Path 101 is the site to go to if you're not sure what you want to do with your life.  When those people click, they're raising their hand and saying, "Yes, that general concept sounds good.. take me to that."  They know a little something about what's being offered and are willing to have you tell them what the site is all about.  They're not "on a mission" to get in and out as quickly as possible with a specific nugget of info or a price or song, etc. 

Similarly, anyone who winds up at Path 101's front door will most likely be typing things into Google like, "How do I figure out what I want to do?" and "Picking a career".  Hopefully, if our SEO is worth a damn, anything more specific like, "Picking a career in finance" would take them to a more specific page within the site.   It will be a while before someone goes to their computer and says, "I want to find a career in finance" , automatically thinks to go to the homepage of Path 101 first, and THEN looks to do a specific search for finance careers.  Basically, we're saying that people either search for specific things at Google or go to your site's homepage because the general concept of what you have interests them, and they're willing to put you in the driver's seat.


Blogged with Flock

Tags: , , ,