You can’t infer intent from screen size

In a recent online discussion about responsive web design versus dedicated mobile sites, I cited this data from a recent study by Google and Neilsen:

77% of mobile searches happen in locations where you would expect a desktop device to also exist—at home and in the office.

In rebuttal, Jonathan Bass shared this perspective:

You are right in that 77% are doing the same. But your implication is wrong, IMHO. The 23% who are doing things differently are precisely why mobile is not the same as desktop. If optimisation is valuable for much smaller segments then it's even more so for a quarter of the audience.

You can't infer intent from screen size

If there’s one mantra I’ve been trumpeting around RWD for the last year, it’s that you can’t infer intent from screen size. Mobile context is a myth.

I see two themes:

Based on the behaviors I observe, I expect mobile traffic for what were previously considered "desktop tasks" to continue to rise. The web is everywhere, and we can't stop it.
We have no way of knowing, based on the size of device, what the person is there to do. We can assume they only want directions or a phone number or what not, but our chances of being wrong are very high.

What we should be doing about it

For me, the direction seems clear:

Adaptive design and progressive enhancement (not content) so that you're serving appropriate and appropriately sized resources based on the device and it's limitations or capabilities.
Fluid layouts that provide the same content to everyone.
A focus on performance as a key design feature—whether the person is on a desktop, smartphone, refrigerator, TV, or whatever.
Really strong information architecture. Identify common use cases and workflows, and make it easy for the visitors themselves to find the content that fits their desires.

It’s arrogant to assume that we as designers and developers can know exactly what a person wants from our site. Our job should be to make it easy for them to get where they want to go.