The UX of Tab Priority

Alot of bad assumptions, especially by non designers on this topic.

The question is: If you have several tabs to use in navigation and you want the user to pay more attention to certain ones, how should you arrange the tabs?

The most obvious one is the left most tab (assuming left-to-right reading) should be the most important section.  However, most people make the mistake right after that.

People assume the second tab is the one that users will read next.  They think people will read the tabs as if they were a sentence in a book.  This is incorrect.  A better metaphor is that the user is a lion.  They are scanning the plains looking for the right animal to leap on.

The user is usually bombarded by things they could click on.  They have mammalian eyes which are accustomed to looking for edge cases.  People tend to see the first, then skim through the middle and see the end.  We are designed to try to make out the shapes of things.  This is why Zebras have stripes.  It protects the animals in the middle.  It’s hard to find their edges.

So if you want the user to click on certain tabs, you should put them where their eyes will naturally go; the left and right edges.  Save the middle for the rest of the stuff.

It’s amazing to me that people will swear up and down that the second tab is the most important.  Unfortunately, this is one of those things that are UX old wives tales.  Sounds good, but in reality are just not true.


Comments

13 responses to “The UX of Tab Priority”

  1. Evidence, please!

    1. Glen Lipka Avatar
      Glen Lipka

      I don’t have access to the tests we did at Intuit, but as an example, look at this heat map image. Notice on the left how the eye starts down, then skips to the end. Or this image from Don’t make me Think which shows the user randomly scanning the page looking for the animal to attack (or link to click).

      There are plenty of studies, but this is not a blog about research. It’s about Design. Read the first blog entry on your own blog about Don Norman. Jobs fired all off the researchers. This tip about tabs is true. But its up to you what to do with it.

  2. Marcus Tucker Avatar
    Marcus Tucker

    I too am sceptical, and while there may be some merit in this claim, much more detail about precisely when it is and isn’t true is needed – with research evidence to back it up.

    For instance, if there are only three tabs and they are located on the left hand side of the page (perhaps occupying only the first 25% of page width) then it’s somewhat plausible that the third might receive more attention than the others.

    However, if there are 10 tabs spanning the entire width of the page then I doubt very much that the rightmost tab will receive more attention than the others – I suspect the opposite, which is the conventional view on this topic.

    Until these (and other) different scenarios are methodically researched and justified conclusions can be drawn, IMHO this bold claim will have to remain unproven hearsay, and therefore not actionable.

  3. Marcus Tucker Avatar
    Marcus Tucker

    Glen – not only is the heatmap image you cite completely irrelevant to the matter at hand, without further context (e.g. is it from a single user or from thousands) it’s hard to interpret, and you certainly can’t derive conclusions about the order in which the areas were looked at. Such flawed justification only serves to further increase my scepticism of your claim about tab priority.

    1. Glen Lipka Avatar
      Glen Lipka

      Any claim based on any amount of research should also be looked at in the context of the specific use case. This is why I included a picture of 4 horizontal tabs.

      The larger point: If you waited for “proof” of every design decision being right, you wouldn’t get anything done. Most products are horrible and most user researchers can’t design a product at all. Do you have proof that the opposite is true? I have done tests exactly like the above picture at Intuit on their public websites and the conclusions were solid.

      Go through the Best of commadot.com at the top and tell me what you learn. A skeptic will learn nothing.

  4. Marcus Tucker Avatar
    Marcus Tucker

    Extraordinary claims require extraordinary evidence – you provide none at all.

    Your image may well include only 4 tabs, but it’s merely an illustration, not a real-world screenshot/application where the tab headings would likely be much longer and probably greater in number too.

    The text of your article also makes no mention of any limitations to your claim, such as that it might only be true for small numbers of tabs or which don’t span the entire page – which is precisely why I raised those specific objections. Instead, your article takes a very black & white stance – stating in the illustration that the second tab is “wrong” and the rightmost is “right”. Similarly, your text states “People assume the second tab is the one that users will read next. They think people will read the tabs as if they were a sentence in a book. This is incorrect.” – you are making sweeping statements which you do not support with evidence.

    There are of course any number of variables which might influence user perception (or “priority” as you call it) of tabs, and the effect of varying them (in isolation and in combination) should be tested before making such claims.

    Following your logic (but at least relying on SOME evidence that I can cite publicly) I could just as easily declare that the first tab should always be coloured red, and the second should be coloured green, based upon the results of http://www.abtests.com/test/97001/homepage-for-performable

    But I won’t, as such a claim would clearly be laughable.

  5. Marcus Tucker Avatar
    Marcus Tucker

    Extraordinary claims require extraordinary evidence, yet you merely assert your claim and challenge me to prove it false!

    Your “picture” is merely an illustration, not a real-world example, and provides no context about the size of the tabs vs the width of the page, etc. The text of your article also doesn’t mention any variables which may affect the application of your claim – it effectively states that it’s true under all circumstances, which is so black-and-white that it cannot be true, and hence another reason that I remain sceptical.

    The onus is on you as the person making a claim which refutes the norms to justify it with evidence. Otherwise it’s just hearsay or wishful thinking.

  6. “It’s amazing to me that people will swear up and down that the second tab is the most important. Unfortunately, this is one of those things that are UX old wives tales. Sounds good, but in reality are just not true.”

    Glen, if you’re going to make such definitive claim, then it’s only fair that you provide some evidence to back this up. Otherwise your opinion is just that, and has no more value than those of anyone else.

    You may well be correct, but why should I take your word for it?

  7. Glen Lipka Avatar
    Glen Lipka

    @Lee, @Marcus: On this blog the vast majority of UX points are based on my experience or testing I have done personally. “Definitive” is not a word I would use much here. I am not a scientist, I am a UX Designer. However, my points are usually stated with passion (and sometimes humor) which could be construed as more exacting that they imply.

    My question to you is: If you have a situation that you can not possibly test. (Like a brand new application that no one uses yet) and some executive says, “Put the [whatever] tab second because it’s more important”. What will you say? What will you think? Will you just assume that he is right? That he is wrong? How would you solve this problem? You can not always test things, that is not the real world.

    Sometimes you have to look around the web and get some feedback. My feedback is “My experience and testing I personally did on public websites at Intuit lead me to the conclusion that the beginning and ends of lists or tabs that people can see on a single screen are the ones that get the most eye contact.” You can take that feedback or not. There is no public information I can see to the contrary so your choices are limited.

    You can be skeptical, but that executive wants a design and you can either say “I’m skeptical and have no other answer” or you do what you think is right.

  8. Hi Glen,
    Could you give some insight into how the tests were done? I have heard your statement by more people and was wondering how you went about testing this. Was it eye-tracking? What other ways would you recommend to test for this?

  9. Glen Lipka Avatar
    Glen Lipka

    @Jeroen: It was quite simple. We had 6 tabs at Intuit.com. We had an A/B/C test. Control was the normal order, Tabs in one alternate order, Tabs in another order. (No other changes) The last tabs had an abnormal bump in click through rates against the control. The bump wasn’t super large, under 5% higher; still it refuted the idea that the second tab would get the bump which is what the “common wisdom” is.

  10. Marcus Tucker Avatar
    Marcus Tucker

    Thanks for adding more detail about your testing, just the kind of information that should have been included in your original post – I hope you’ll do so in future?

    However, given that it’s limited to only a single test case (i.e. 6 tabs at Intuit.com) and it resulted in less than 5% difference, I think extrapolating your findings from that single specific case to reach a general conclusion about all tabs is unscientific to say the least – e.g. “It’s amazing to me that people will swear up and down that the second tab is the most important. Unfortunately, this is one of those things that are UX old wives tales. Sounds good, but in reality are just not true.” – contrary to what you claim, your single test does not in any way prove that conventional wisdom regarding tab perception (i.e. left to right) is false.

    That said, it’s certainly something that might be worth investigating further, and if testing across a wide range of scenarios in which tabs are used reveals a general trend supporting your theory then you may in time be justified in making such claims, and would be able to cite data to support the case that you make. But not yet!

    Finally, with regard to the heat map in your first comment/response, from which you claimed it was possible to infer the order in which things were looked at, I highly recommend this excellent presentation on eye tracking technology and how heat maps can easily be misinterpreted, which was delivered at the recent UX Lx conference in Lisbon (hat tip: UX Booth).

    1. Glen Lipka Avatar
      Glen Lipka

      I can’t make conclusions? or shouldn’t? If I waited for scientific proof, I would have to just accept the executives own UNPROVEN theory that the opposite is true. UX doesn’t have proof of everything. Sometimes you need to make a stand and use a hypothesis. Your points are valid but don’t help a UX designer make a decision. We can’t test everything.

Whatya think?