This was originally a comment to the article titled Web Second, Mobile First on Mark Suster’s excellent blog Both Sides of the Table. While Mark has a consistently high level of quality in his articles I was, to be honest, a bit disappointed with this one. First of all, it states a number of observations that seem pretty obvious to me (and therefore, I believe, to most everyone), such as that a smartphone (or “mobile”) is increasingly the first computing device for many new users.
But more importantly, it’s continuing the false dichotomy of “mobile vs. Web”. Why it’s false? Simply because modern mobile devices — at least those with their own ecosystems — are perfectly able to display Web, and it’s becoming extremely easy to develop for both mobile and Web at the same time, with only a few more resources devoted to ensuring the cross-platform compatibility (which are necessary even if you develop only for “Web”, as you need to take into account different browsers and OSs: e.g. if you’re aiming at China, 77% of your visitors will use IE 8.0 and earlier).
If you’re strapped for resources, there is absolutely no need to develop a mobile app and have to depend on the whims of AppStore and other walled gardens out there — you can develop for the Web, and make your front-end switch the styles automatically according to the device it’s viewed on. You say that you love the new LinkedIn mobile app; but have you seen their mobile Web? It’s pretty much as functional as the mobile app, looks just as well, and has probably required only a bit more resources than the “traditional” Web app.
Essentially, in my opinion, a mobile app makes sense only if it requires no Internet connection to operate properly, so it’s perfect for games, fart jokes and similar use cases. For all the examples Mark mentioned in his article — Yelp, LinkedIn, Foursquare etc — ubiquitous Internet is a prerequisite, which means that there is no advantage over a mobile Web app. Actually, with the modern HTML5 features such as local storage, even a less than 100% reliable connection is not necessarily a problem, as some data can be stored locally and used when offline, syncing it back to the server when online.
So, instead of the “Mobile first, Web second” approach, I’d suggest a different strategy to most new startups: “Web (classic and mobile) first, mobile perhaps (if necessary)”.