American is not a default setting (or how to build truly global products)

Lessons learned from 7 years running a global team, part 1

Isabelle Roughol
Isabelle Roughol

I knew I’d made a difference when product managers started preempting my objections with a laugh. “Here’s the slide that will answer Isa’s question,” they’d say in presentations. It was an earworm I had dropped so many times it became my accidental catchphrase: “How does this work internationally?”

I spent much of my seven years at LinkedIn expanding our editorial team and products globally. They now serve people in 9 languages and nearly 100 countries. I learned a lot from what we got right, what we got wrong, and what we got right eventually. Here are a few tidbits I thought worth passing on.

Because the lessons were many, this first piece is guidance for those in leadership and at company headquarters. Another piece will follow for professionals in the trenches building international outposts. Please share your own experience and advice in the comments.

  • American is not a default setting

I’m not picking on the US here, but Silicon Valley is what I know. Companies from anywhere, especially when they have a large home market, behave the same way. When they grow “internationally,” they’ll often think of their home users as the default and of users abroad as those that veer from the norm. Products must be “adapted.” They are built, then “internationalized” – if there’s time. Yet, international only means “of several nations,” not “everywhere but here.” If you’re to build a digital product that is truly for all, you can’t just slap subtitles on it. Home is a “geo” too and we are all someone’s foreigner. Every product version, every customer, every user is simply a variation on a theme. It’s a small mindset change, but it makes a world of difference. From it will flow the right language. You’ll never again use “rest of world” in an org chart.

  • Diversify your team

You don’t know what you don’t know. You’ll miss a lot if your employees do not share the lived experiences of your user base. Employ native or equivalent speakers; make sure they have language skills but also cultural knowledge. They’ll save you from embarrassing (and often offensive) mistakes. Ideally they should be in the country, and not just of it: speaking as a longtime immigrant, it’s easy to quickly lose touch with the cultural zeitgeist of home. Value globally minded employees who can code switch between cultures and build the necessary bridges.

  • Incentivize global work

Much in Silicon Valley encourages technologists to build products for their friends and the people who will hire them next. That’s how you get a hundred apps helping wealthy urbanites farm out their life admin to gig workers. Make sure building for a different kind of user is rewarded. Don’t celebrate early launches but completed global ramps. (Google calls them product landings, it’s cheesy but whatever works for you…) Don’t move the team on until their product functions everywhere. Be generous with the travel budget. Assign senior people to foreign markets and design accelerated career paths for those working internationally. Make global mobility easy. Do anything that will get a Stanford grad excited to build for a Kenyan banker or German farmer, and make sure there is glory in that work.

  • Build for the future

It may take a bit longer but pays off to build products that will scale globally from day one, rather than have to retrofit later. Ask any engineer who’s seen their platform brought asunder by a special character or who’s had to adapt it to bidirectional writing. Clean code that can scale is a matter of craftsmanship; LinkedIn engineers had an award for it. Think of it as building foundations that will withstand the house extension when the third kid comes. You don’t have to go global right away, but you’ll be glad not to have to tear down the house when you do.

  • One country isn’t a yardstick for anything but one country.

Bad sampling will ruin the best experiments. Building a global strategy off local metrics is the equivalent of testing a drug on mice and immediately putting it on the market for humans. That’s just bad science. Testing everywhere isn’t practical, but pick a few significantly different markets for user research and A/B testing before declaring a new product a success. Also consider ramping products in other countries first, rather than always at home.

  • Walk a mile in your users’ shoes.

LinkedIn ran 2G Tuesdays: you could opt in to throttle your phone download speed for a day and experience the app the way millions do every day in regions with poor connectivity. It was an eye opener. The international engineering team encouraged employees to use the app in whatever foreign language they took for a couple years in high school. Such policies help give the whole company awareness of foreign users, even if it seems remote from their work.

  • Loosen the reins on faraway colleagues

Running a global team isn’t for micromanagers. You can’t tell whether people are at their desks. Heck, you’re lucky if there’s an hour in the day when your time zones coincide and you can catch up. In-market employees are likely to know better than you do how their country works and what their users need. Align on goals, and let them get there their own way. Judge on results.

It’s for those faraway people I will write the next post. Stay tuned. Should I dig deeper into any of these points? What other lessons from building global teams would you share? Let me know.

Above: Shanghai by night on a 2015 LinkedIn study trip to China. (Photo by Isabelle Roughol)


Isabelle Roughol

Journalist. Founder & host of Borderline. Former international editor of LinkedIn, foreign editor at Le Figaro, reporter at The Cambodia Daily. Global soul, messy accent.

Find the podcast & listen now

Find the podcast & listen now