Wrong and poor advice to subvert browser functionality. Alert the user to which input they are currently working with. Eliminates [tiresome] user guesswork and is conventional. Also helps the visually challenged.
> via overlay boxes
I don't even.
> I enter my city and hit ‘Enter’. That triggers the search, without giving me a chance to enter my dates first.
It occurs to me that this is part of the design. They show you the inventory before asking you for your size, even though if you had the patience you can input your desired dates on the homepage. Also, note the date input is also available after form submission. Something tells me that they've done some research on user behavior.
> How else can Airbnb improve?
I'm going to stop now because if anything, this post is novice-level and poorly thought out if one is assuming that they can make better what they don't understand.
Agreed it should outline per browser functionality but you can fix this by placing the button inside the input and highlight the entire box.
I don't think lightboxes are bad for login especially if the user is already in the middle of a task. Different case if the site requires a login to function correctly like Facebook.
I've been thinking through this on our hotel search engine as well. We've decided to place the search box below the date inputs because dates are integral to pricing which is a major component of the shopping decision making process. For AirBnB, it looks like prices for properties are relatively fixed so it's not as big of an issue since they know the pricing even w/o dates that are right most of the time. The issue I have is availability. If I don't enter dates, would much of what I'm seeing actually be irrelevant to my actual travel itinerary if they are actually not even available?
Additionally, the ranking of results when you search is processed by Google places autocomplete. The request never actually hits any AirBnB server for ranking. This is done on Google side via their friendly API.
Subverting browser behavior happens frequently and can be a positive thing. Lightboxes achieve a UX experience that bouncing someone elsewhere does not.
I personally agree with using light boxes, IF they are done right. If the connection speed and browser can easily handle it, why not? But still always having the permalink "/sign-up" (or whatever it is) so you can link to it.
Can you provide more insight on why you are against that? Just curious if I missed the boat on people hating light boxes now? haha
There's usually no indication within the content that what you're clicking on, usually a link, is in fact not going to take you to a separate URL where you can complete your business, which is conventional, but it in fact going to change your visual experience by [poorly] applying an opaque [non-universal] overlay to your content and present you with a form, content, media in a format that is completely unnecessary to the experience. In the case of the author's preferred third-party plugin, it adds an additional four steps to your experience:
1) activate overlay, 2) resize viewport to content, 3) present content, 4) user-activate close button.
All for content that is preloaded with the page. The author seriously thinks that inconveniencing the user with added steps and amateurish functionality at the moment of conversion is a good idea, but you can browse the code over at their webpage to get a feel for what they think is acceptable advice.
The author-suggested solution takes four to five times longer without even making a server call. If that isn't ignorant, I don't know what is.
If look at Google.com, the outline is a single pixel color stroke that replaces the gray stroke that currently exists when it's not in use.
When you look at Facebook.com, they don't have a highlight either.
Maybe it's my own view, but when a designer circumvents or subverts basic, conventional browser functionality it signals they don't know how to design for the web or they care more for their particular design than they do for the user's ability to process and react to the intended functionality.