Skip to main content

Don’t Kill Yourself: Mobile Web Performance Tricks that Aren’t Worth It—and Some that Are

Lyza Gardner (Cloud Four)
Location: 113 Level: Intermediate
Average rating: ***..
(3.63, 30 ratings)

Pain for gain. Certainly—sometimes. Meaningful leaps in performance occasionally require heroic effort. But is it always worth it? And does it have to be so hard? There’s quite a collection of RWD and mobile web performance tips and hacks out there; let’s assess which ones deliver satisfaction and which ones leave you sweaty, angry and with little real result. The proliferation of RWD and mobile sites is giving us enough data to emerge from theoretical analysis out into the real world of measurable statistics. Out there in the jungle of real numbers, we’ll find some counter-intuitive surprises that may leave you baffled but smiling.

Finding reliable performance touchstones is challenging in a technology landscape that has unreliable consistency. Even basic performance 101 tenets, like reducing requests, have variable outcomes in real life.

How much can we, as implementors, really adapt before the process fails to scale? Changes on the horizon, like HTTP2/SPDY, may force us to revisit tactics like domain sharding. Is there a way to distribute the burden of performance complexity so it doesn’t drive us completely mad?

And, really, is there some point at which we can draw a line and say: “Enough! This performs well enough!” without feeling like we’ve abandoned ship?

Photo of Lyza Gardner

Lyza Gardner

Cloud Four

Lyza Danger Gardner is a dev. Since co-founding Portland, Ore.-based mobile web start-up Cloud Four ( in 2007, Lyza has tortured and thrilled herself with the intricate ins and outs of the bazillion devices and browsers now accessing the web globally. Lyza and co-founder Jason Grigsby are the authors of Head First Mobile Web (O’Reilly).

Comments on this page are now closed.


Andrea Maurizzi
25-11-2014 18:24 CET

Since I really appreciated the speech, is there a chance to have the slides and/or the recording of it? Thanks.