A lot of my job is done abroad. This year I spent almost two and a half months abroad, 73 days all told. Being out of the States so often and for so long, cumulatively, gives me many opprotunities to learn and to remember things I’ve forgotten since moving back to the US in 2008. I really appreciate these chances, even if some of them are lonely, or represent significant challenges at work. Enough are interesting and for personal adventure to keep me happy, and keep me traveling.

2014 brought one specific change to my travel methods, and because of that an experience I wanted to share. I no longer use local SIMs, save in extraordinary situations. In October of 2013, T-Mobile, an American mobile phone company, launched free international data roaming. Even now, more than a year later, typing those words feels amazing. Free international data. To give context, previous international data deals available in the US ran something in the realm of $30 USD for 50 megabytes of international roaming data. Thirty dollars for fifty megabytes. It’s easy to see why I switched to T-Mobile.

The catch, because of course there is one, is that this free and unlimited data comes down from the tower at 2G speeds.

So I spent one fifth of 2014 on 2G, and the remaining four fifths on LTE. Or with no service in the wilder parts of the US, specifically northern California, north western Colorado, and a lot of the cross-country train ride. That is another trade-off that comes with chosing T-Mobile. It’s an easy choice for me, being primarily a city person.

Having free international data and spending so much time on the road, be it in the trains of Hong Kong, Shanghai, and Tokyo, or the traffic of Manila and Shenzhen, not to mention factories, restaurants, and hotels, means a lot of phone time. A lot of email. A lot of Twitter. A lot of web. And that leads to the point. In 2014, the web is hard on 2G. Sites load slowly, first displaying banner ads and only then, tens of seconds later, the all-text content of the article. Mastheads take dozens of seconds to load, complex drop down menus and high-resolution logos. Analytics packages. And ads. Some activities and apps simply don’t scale well to 2G. Instagram, for example, is an exercise in patience, but a worthwhile one. And Google Maps, well…

In 2014 it feels like the network is finally everywhere, or almost. And it feels like the future. Being able to turn on my phone in any country on landing and check on my cat, at home in San Francisco, will probably still feel surprisingly wonderful for another couple of years. And 2G isn’t bad for most things. Despite how it probably sounds, this post is not meant as a complaint. It’s meant as a note, a reminder, and a future consideration. For example, loading time for maps matter. More than anything, maps are used when in unfamiliar locations, and often those are situations without great network access. Be it hotel wifi, 2G cell network, or just the slow connections of many smaller cities, maps are most necessary on the fringes, out of our comfort zone, and often in something of a hurry. Yes, most of the places I’ve been have faster networks. Hong Kong has excellent service, faster than the US in many cases. But not every place does. Not every city has LTE, nor every carrier, and that’s the point.

I view these 73 days on 2G as a test of how we interact with networks, and as a challenge for service design. Twitter as it used to be, all text 140 characters or less, was the perfect low-bandwidth mobile-first service. Modern Twitter, with video, photos, expanded links, and soundcloud embedded, is increasingly something built for fast networks, for always-on connections. Not necessarily a bad set of decisions, but a definite shift from a service originally built on SMS, built for the mobile networks we used to have and that many still do.

Of course not all things are built for slow mobile networks, and that’s fine. Heck, Tumblr is one of them, image heavy and full of .gifs. Oh god, .gifs on 2G. If ever a format’s resurgence has come without consideration of bandwith, .gif is it.

Overall I think a few weeks on 2G is something product teams should experience, and consider, not just today or this year, but well into the future. There will be people on slower networks and with worse connections for much longer than San Francisco, which had quite poor cell networks just a handfull of years ago. If a service is designed to “change the world” it needs to be usable out in that world.