Why Left Search Buttons Perform Faster Than Right

16 Nov 2010 - 11:06pm
3 years ago
72 replies
3961 reads
uxmovement
2010

Research has shown that having or not having a search button makes no difference to pro users. However, for novice users, right search buttons are generally slow and inefficient. To me, the solution is simple. A left search button reduces visual fixations, puts the label first and makes doing searches faster for users. 

The pattern left search buttons follow is similar to IXDA's search where the combo box and icon come before the text field. The label and button have more visual weight and are what users need to see first before they carry out a search, which is why I believe they should come before the text field. Despite the fact that most websites and search engines have it backwards, what are your professional and experienced independent thoughts on this matter?

http://uxmovement.com/design-articles/left-search-buttons-perform-faster-than-right for further detail.

 

 

Comments

20 Dec 2010 - 12:58am
johnk
2010

I don't want to argue about this - but there are some users who use the mouse to copy and paste text.  Those people need some kind of button to substitute for the Enter key to submit the form.

If it's too obtrusive to have a button on the right, perhaps it can be designed to work like the Firefox address bar.  In FF, you paste the URL, and then a green arrow appears in the address bar - you click it and go to the URL.

20 Dec 2010 - 10:03am
Dave Malouf
2005

As for the handle features, can't we just easily change the field where people sign up to First/las name instead of "Username" and then just have people log-in by email only please?

I also think people should look at Quora. I'm not in love w/ it, but it definitely has some interesting aspects.

Dan while "knowledge" is 1 aspect of IxDA, community is another and I don't think that the Quora model handles community in their solution. Maybe it speaks to a bifurcation of tools that is needed to address both, but that would be a huge change in the community dynamic of this virtual community. Do you have thoughts on how to address both of these components?

-- dave

20 Dec 2010 - 3:36pm
Paul Bryan
2008

On 29 Nov 2010 - 7:05pm, Dan Saffer wrote:

> The validity of opinions rests on authority and reputation.

Not so. Authority depends on reputation. Validity depends on evidence. The idea that validity is related to reputation in any meaningful way is a logical fallacy. "Appeal to authority" seeks to persuade by citing what someone else, a perceived authority, thinks on the subject, as if that resolves the question.

The reason I would pay more attention if Jared Spool or other authority says something unexpected (e.g. about remote usability testing), is that I assume he has evidence that I don't know about yet. The validity of what he says, however, depends on the evidence, not how well known he is.

 

Paul Bryan

http://www.virtualfloorspace.com

 

20 Dec 2010 - 9:05pm
Benjamin Barnett
2009

In context of Innovation flow...

Why is observation / experience not considered a potential determining factor around these parts? Data in concept enviro's is subjected to the parameters of your (theorized) models and can be manipulated, exposed, or just be random as well as added as a key factor (more experience) after experimenting. I suppose you can /should produce (show it and prove it) data to justify your actions, but experience goes a long way in the early moments of innovation flow.

Besides, ask NASA and any Biologist if more than 6 elements could exist 1 month ago. They would have said "not possible" and laughed.

They and the history that proceeded them were flat out wrong.

So much for the Data.

Sent from my iPhone

On Dec 20, 2010, at 4:06 PM, Paul Bryan wrote:

> On 29 Nov 2010 - 7:05pm, Dan Saffer wrote: > > > The validity of opinions rests on authority and reputation. > > Not so. Authority depends on reputation. Validity depends on evidence. The idea > that validity is related to reputation in any meaningful way is a logical > fallacy. "Appeal to authority" seeks to persuade by citing what > someone else, a perceived authority, thinks on the subject, as if that resolves > the question. > > The reason I would pay more attention if Jared Spool or other authority > says something unexpected (e.g. about > remote usability testing), is that I assume he has evidence that I don't know > about yet. The validity of what he says, however, depends on the evidence, not > how well known he is. > >
> > Paul Bryan > > http://www.virtualfloorspace.com > >
> >

20 Dec 2010 - 9:05pm
Dan Saffer
2003

On Dec 20, 2010, at 1:04 PM, Paul Bryan wrote:

> On 29 Nov 2010 - 7:05pm, Dan Saffer wrote: > > > The validity of opinions rests on authority and reputation. > > Not so. Authority depends on reputation. Validity depends on evidence. The idea > that validity is related to reputation in any meaningful way is a logical > fallacy. "Appeal to authority" seeks to persuade by citing what > someone else, a perceived authority, thinks on the subject, as if that resolves > the question.

I should have said LIKELY validity. Obviously proven facts trump opinions, but without those, and all other things being equal, I would trust the opinion of an expert in the field over someone without because his/her opinion is more likely to be proved correct over time, via evidence.

Dan

20 Dec 2010 - 9:58pm
Jared M. Spool
2003

Thanks Chief,

You just save me an awful lot of research work.

:)

Jared

21 Dec 2010 - 6:06am
Alan James Salmoni
2008

Saved? I'm not sure I like the idea of testing the entire world's population (living, dead, and yet-to-be-living) to find anything out. Thanks but I'll stick with sampling for now until Wowbagger the Infinitely Prolonged agrees to help me out.

AJS
On Tue, Dec 21, 2010 at 4:09 AM, Jared M. Spool <jspool@uie.com> wrote:

Thanks Chief,

You just save me an awful lot of research work.

:)

Jared

20 Dec 2010 - 11:46pm
Floris Kruger
2010

Very well spoken, Dan. The point needn't be argued any further, in my opinion, because your last paragraph thoroughly covers all bases.

Also, while I said you had a condescending tone, I should actually commend you for not having unleashed your initial draft upon poor uxmovement - who appears to have disappeared from his thread.

Floris

4 Jan 2011 - 3:03am
Floris Kruger
2010

Paul Bryan, you nailed it.

21 Dec 2010 - 4:47am
Floris Kruger
2010

Google currently sets the standard in Search logic and efficiency. With respect, the only search box I am interested in discussing at this time is the one that improves upon the Google Search experience, because theirs is the one to beat.

I think it's great and very important for you to question everything, uxmovement, but go back to the drawing board to find indisputable rationale for your ideas. As Dan mentioned, that's how your opinion can gain credibility. Plus, the mere act of attempting to prove your ideas can actually resolve them for you. If you can't prove an idea, just pose it as a question, and you'll find that others are much more receptive to jump in and offer clarity.

21 Dec 2010 - 9:11am
Jared M. Spool
2003

Florian said:

Google currently sets the standard in Search logic and efficiency. With respect, the only search box I am interested in discussing at this time is the one that improves upon the Google Search experience, because theirs is the one to beat.

Really? Google's search experience works great when searching across billions of web site pages where page rank works well.

Their experience is truly substandard for on-site search, especially for any specialized content, such as product searches, citation searches, clinical trial searches, or a raft of others.

I would think it would be very easy to improve on the Google Search experience for 99% of things designers need their users to search on.

That's my opinion. It has all the validity and authority that comes with me rolling out of bed.

Jared

21 Dec 2010 - 11:01am
Paul Bryan
2008

Authority, not validity  ; )

22 Dec 2010 - 5:14pm
Floris Kruger
2010

Jared, I see you just made a grand entrance in my Inbox via the upcoming, Aquent sponsored, UIE Book Club event on Jan 4. Much respect, sir.

21 Dec 2010 - 5:07pm
Floris Kruger
2010

Jared, in keeping w/ the topic, I'm referring to Google's search box treatment. However, I have no doubt that Google would also own the on-site search experience if that was actually a priority of theirs. And if their intention was to create a search experience for designers, they would ace that too. Don't think for a moment they haven't already considered what you're referring to.

 

21 Dec 2010 - 5:55pm
Floris Kruger
2010

I love to be part of this community, but gosh, moderation is very slow. I don't know how long I can keep communicating at this pace. Ideas happen so fast and we make such little headway at a time.

21 Dec 2010 - 8:06pm
Floris Kruger
2010

Whew... that's it for me, folks. I can't wait a whole day for a conversation to be advanced by one paragraph. Maybe the site owners should migrate IxDA to a Google blog :-) Peace out!

22 Dec 2010 - 4:56am
Floris Kruger
2010

Kindly disregard my last comment. Moderation lifted. Thanks so much you guys! This is much, much better.

23 Dec 2010 - 5:59am
shetty
2010

General logical flow is left to right except in middle east countries, I feel search text box + Search button is the better option for the user.

29 Dec 2010 - 9:44pm
Samantha LeVan
2009

I've really debated whether I want to join this discussion. As someone who has facilitated more than 500 usability test sessions and observed plenty of others, studied search workflows for eCommerce and intranet, and super inspired by the many wonderful experts involved with IxDA, I was frustrated at this thread. I believe that solid, properly researched data needs to support a major change in a web standard like this and I don't see that has happened. It is unfortunate that someone with an interesting hypothesis presented the idea as he would a proven fact when it could have been an awesome change for talented researchers and designers to collaborate to prove something either way.

I have a hard time with the pro vs novice concept in search usage. I've seen many people who use the internet daily and have for years, but still choose to use the mouse for actions and keyboard for data input. Those are people who may be experts at search and internet use but won't use the enter key. For them, we can predict their task time using CogTool (cognitive modeling using KLM-GOMS). I did a quick eval using Yahoo's search and found only a .004 difference in time. However, this doesn't factor in the amazing amount of confusion I would predict if we asked people to think right-to-left instead of left-to-right, and that is why you can't just rely on one small piece of data in these situations.

I hate to see someone ragged on so much but I hate just as much to think the search button could end up on the left of a search field.

4 Jan 2011 - 2:08am
Floris Kruger
2010

We have inadvertently pushed this thread to the top of the Hot Discussions list.

5 Jan 2011 - 1:30pm
nixkuroi
2010

For what it's worth, while the visual fixation problem is there, I'm not sure it outweighs the expected navigation problem it creates.  In any application with form elements (word, acrobat, excel, default html behavior, etc):

Click the first text box, hit tab. 

Repeat until the bottom and tab to the submit button, then enter.

The left side button breaks this navigation standard.  If you're wondering if it's really a standard, you have only to look at the tabindex attribute which continues into every version of any application layout engine I've ever encountered.  

I think you really have to ask yourself if this behavior is limited to pro users or not.  I'm inclined to believe that even an office worker who's been working in Windows 3.1 or Mac Classic for twenty years and gets on the internet tomorrow is going to expect this behavior to continue.  

Thoughts?

5 Jan 2011 - 3:26pm
Floris Kruger
2010

Very good point. It would mess up the tab index, yes, which is an accessibility issue for both keyboard use and screen readers. Doesn't mean it can't ever be changed, though - just so long as any new paradigm also adequately accommodates accessibility.

Syndicate content Get the feed