Dropbox’s Playbook for International Expansion, with ChenLi Wang

Special thanks to Chenli Wang for his participation in this post.

There have been two pillars to growth at Dropbox. The first, Dropbox’ freemium model based on word-of-mouth referrals, is something that most of us have experienced if we count ourselves among the company’s 500 million users. The second driver behind Dropbox’s growth to hit the 500 million user milestone is something that’s less visible many of its core users.

As of this year, 75% of Dropbox’s users are now outside of the U.S. In the past year and a half, Dropbox finalized key distribution partnerships with Softbank, Vodafone, Telmex and other partners that help carry it into new geo-linguistic markets (as well as onto different platforms). 

We recently spoke to ChenLi Wang, who oversaw international expansion as head of the company’s product and business operations teams, about frameworks for international growth, which countries and what timing, deciding on your first international outpost, some unexpected benefits of international, and its biggest risks.  

Table of Contents

1. A universal framework for international growth?
2. Which countries, and when
3. When is the right time to do international growth?
4. How to decide on your first international outpost: Don’t just copy Google
5. What happens when you actually open an office
6. Your initial landing team
7. International might not be as different as you think
8. Or international could be really different. How to run growth experiments in a completely different “lab”
9. Data risks and regulatory norms
10. Currency and payments challenges with international expansion
11. Biggest mistakes in international launches

Get our growth essays delivered, one email per post. 

Dropbox’s Timeline of International Expansion
Image courtesy of Qordoba 

Wang had worked on international efforts at previous companies, but saw a unique spark of momentum in Dropbox from his first days there. In an interconnected world, a company that’s already spreading virally from person to person even before any intentional international strategy will begin its international growth on its own. 

Whether or not organic international growth might happen for a given product depends on the demographics of its user base and how connected they themselves are across borders, Wang says. Dropbox started in 2007, and the product was not localized into any additional languages beyond English until 2011. But, by the time that first round of additional language versions launched, the international user base was already substantial. 

“It was somewhere in the 30% to 40% range already at that time. The user base itself was international, and so in some ways our localization strategy and plan was a “Follow the users” approach as opposed to a “Create the growth from scratch” approach.” 

Dropbox relied on a user-driven and virality-based approach to growth, rather than a strategy driven by sales or marketing. Companies that use on these latter methods of growth might have a harder time easing into an international strategy with significant internal reorganization required to focus on and succeed in expanding their business internationally, says Wang.

“It’s definitely more the exception than the rule. In my previous role at Workday, which uses a sales-driven, top-down enterprise strategy, international was a much more about a disciplined, planned growth approach. It wasn’t already there, waiting for us to do something more with it. 
In fact, the Workday product itself, which includes HR, payroll, and financials, must be tailored to comply with the laws and regulations of different countries.  Dropbox the product has the benefit of being pretty homogeneous and universal.”

On the spectrum of enterprise to consumer-facing products, you'll generally see characteristics of less regulation, more peer-to-peer and word of mouth distribution, and thus more organic international reach towards the consumer end.  This is an important consideration in deciding when and how much to invest in international.

A universal framework for international growth?

With so much variation by business model and product type, can you still apply a broader framework for international growth? Can such a thing as an international growth playbook exist? 

Wondering the same thing, Wang says he consulted with peers at other companies like Pinterest before diving into international expansion at Dropbox. Despite Pinterest’s seemingly universal appeal and its consumer-facing model, his conversations revealed some surprising challenges. 

“They didn’t necessarily see as much of the natural, organic expansion internationally as we did. They’ve done a good amount of user research, and have seeded content in some countries to help people to understand what the product is all about. They’ve done a great job and it’s working for them.
I believe this is more of what most companies should expect to face, and that in our case at Dropbox, it was really an exception rather than the rule.”

As for whether there’s a general framework for international expansion, Wang agrees that there are some basic strategic questions that most companies should ask. Once a business has reached product-market fit and is starting to think about their next horizons of growth, that’s the right time to ask whether or not they should tailor the product for an adjacent second market. That next frontier could be international, or it could be a different platform, a second product, or a different customer base. 

Sample evaluation of international expansion against feature launch, using the ICE framework.

Sample evaluation of international expansion against feature launch, using the ICE framework.

Wang believes that all of those expansion options should be evaluated against similar criteria: 

Given the characteristics and opportunity sizes of your product and market, and the capabilities of your team, what next direction makes the most sense?  

Often times, international won’t rank at the top as compared to others on ease of execution and size of opportunity. Getting international right (ease of execution) requires significant time, internal bandwidth and specific know-how. The size of the opportunity is hard to evaluate when compared to looking at an adjacent market in a country you're familiar with or looking at an adjacent product that you can expand into, if a competitor is already doing it. 

Which countries, and when? 

How much do geographies matter when considering international growth? Is it smarter to pick countries that are more similar to your current dominant markets, or look for more distant opportunities?

“A lot of the early discussions we had at Dropbox about overall growth strategy were around an almost rhetorical question about what’s the number of users you could have by geography. At Dropbox we used some mix of broadband and mobile user numbers (publicly available data you can find online) to give us a rough sense of the opportunity in a given market.”

Wang and his team broke down the results of their thought exercise into two categories of countries. The first group were countries where Dropbox was already doing well, had some amount of brand awareness, and would need to follow a double-down strategy to continue growing there.

The second group were countries where Dropbox was barely present, reaching approximately 1% of the addressable country market. If the company were to spark growth and take that from 1 to 10, it would represent tremendous growth. 

In the first group of countries — Dropbox’s relatively mature markets — the ceiling wasn’t too far off. The people who wanted their product had probably already found it, and the ones that hadn’t already been captured were probably not easily capture-able. 

“Our two quintessential examples were Thailand and the Nordics, particularly Denmark, Sweden and Norway. At that time, Thailand and the Nordic countries added up were approximately the same size in terms of total opportunity.
In Thailand, we were very small but growing very fast. In the Nordics, Dropbox had strong penetration already, and growth was slower.”
In the end, we realized that the ‘double down saturation’ strategy was more viable than the ‘spark growth’ strategy.”

Even in mature markets, Wang’s team found, there were usually addressable reasons why people hadn’t yet adopted or didn't use the product as intensively. Incremental lack of awareness and other sticking points were remaining issues that the team could work on to incrementally grow the product’s reach and deepen its engagement, while continuing to reap the benefit of network effects.

In low-penetration markets, on the other hand, there was usually a really good reason why the product wasn’t doing better. High initial growth off a small base will often start to trail off, and the reasons become clear later.  

“Our Dropbox-specific learning was that while many people found the product useful — it helped them to be more productive etc — ultimately your value as a tool or a piece of software is directly tied to the value that your end users create. And, the value you can charge for your product is a percentage of the value that they create. 
Effectively, this means that your market size is a percentage of the GDP per capita of your end users. Once we applied this framework and looked at all the data, everything fell into place. 
We learned that while usage was universal, monetization (for us) was very much correlated with GDP per capita.” 
Thailand shows an impressive growth curve, but Norway represents greater total opportunity.

Thailand shows an impressive growth curve, but Norway represents greater total opportunity.

Get our insights on international growth, one email per post.

When is the right time to do international growth?

At Dropbox, international growth was happening organically even before product localization, which posed an opportunity and a challenge for deciding if and when the company should pursue a more structured and dedicated international growth strategy. 

“On the one hand, we were saying, ‘Wow look how great we're doing on international, we should do something!" and on the other hand we were also saying, "Wow, look how great we're doing on international, why should we do anything!" 

In the end, Dropbox’s users helped the team get to its answer. The company has prided itself on its customer support from the beginning, and for its premium tiers, technical support is an inseparable part of the core product experience, and an important differentiator from its competitors.

“It's not just about the product working. It's the fact that we offer excellent support to help you solve your problems.” 

It was on this key point of differentiation that Dropbox’s organic international growth started breaking down. or a company based in Silicon Valley, the biggest hurdle was providing technical support to users in countries like Japan and German (two of the largest non-English markets) in their language and in their time zone. The pool of prospective hires that were highly qualified in providing technical support and had German/Japanese language fluency was basically zero in San Francisco.

“We relied on Google translate. There was a built-in plugin through Zendesk where you could do that. But, that’s obviously a sub-optimal solution and anyway it's one you can only do via email. 
As we started expanding into B2B, we were offering phone support. That’s where it started to break down. If you can't actually speak on the phone with your B2B customers in Germany and Japan, then it’s not going to work.” 

Dropbox’s forcing factor for going international ended up being the lack of specific talent and expertise to deliver their existing service.

How to decide on your first international outpost: Don’t just copy Google

Dropbox’ support challenge led it to open its first non-U.S. office in Dublin, Ireland — and though it may seem obvious now, it was a carefully-researched, heavily-debated, and ultimately data-driven decision. 

One of Wang’s team members was in charge of determining in which city the company should base its European headquarters. 

“Some of the locations we were evaluating were Dublin, London, Amsterdam and Zurich, because those were the cities where a lot of other tech companies happen to have substantial presence. 
To decide, we ran an experiment. First, we went to the government-sponsored development agencies and a few local headhunting firms in each location and asked them for stacks of profiles for sales and support roles, with the identity info blanked out.  
Then, we handed these CVs to our sales and support hiring managers and asked them to rank the top candidates, without telling them which countries they were in.
After they went through about a hundred of these, it became very clear that the talent in Dublin was just head and shoulders above the other three cities, for the roles that were important to us.” 

Wang believes that Dublin’s strength in sales and support has been driven by Google and its extensive operation there, but doesn’t think companies should just go where Google goes.

It’s important to understand what particular talent and international goals your company is after. If you’re looking for brand advertising sales or outbound sales, then London or Paris might be a better bet. If you’re looking to build out an engineering team that can drive large complex technical projects, then Zurich should definitely be high on the consideration list (Google has a huge engineering presence there). If it’s design talent you’re after, then Amsterdam would be a very compelling choice. 

“You really have to more thoughtful than just, ‘We're going to copy Google. If you’re going to copy, then at least ask yourself what part of Google do you really want to copy? Look at the part of their business that is relevant to your opportunity right now.”

What happens when you actually open an office

When Dropbox first started its structured international expansion, another experiment it ran was offering its live chat sales service during European working hours (while the landing team was still based in San Francisco). 

The initial results weren’t impressive. The team saw a ten percent lift in incremental sales, which was meaningful, but not the kind of resounding “Yes” that a costly and effortful international expansion effort should be based on.

And yet, six months later, when Wang was in Dublin visiting their initial landing team, he heard a different story. 

“We were sitting outside our office enjoying lunch, and they said, ‘Wow, it makes a huge difference to be here.’ I asked them what they meant, because you know, we had run those experiments and saw that being on chat during European hours didn't really change things that dramatically. 
My team member said, ‘You know when we first got here, that's what it felt like. But things started to pick up after a lot of local PR (publicizing Dropbox’s office opening there), and people starting to notice that we actually have an office here.’” 

Wang says that Dropbox’ Dublin team members were frequently asked by customers where they were based. When they would reply, “Europe” or “Dublin,” there would be a noticeable warming, a drop of the other side’s guard. This wasn’t yet another foreign company doing its global thing; this company was "local."

For Wang, this was the biggest surprise of all. 

“We were a data-driven organization, and yet, we couldn't have run an experiment to find that out. We actually had to go there; you can't test or run an experiment for everything. 
That was an unexpected surprise that ended up driving a lot more growth than we had forecasted.”

Your initial landing team

When Dropbox opened its initial office in Dublin, it wasn’t just a matter of hiring local staff to carry the product, and the company’s culture, to the new region.

Instead, they sent core staff from Dropbox SF to live and work on the ground for a year and a half. Landing team members weren’t specifically chartered with “international growth,” however. Instead, it was as if they were running a startup within the company, and that startup happened to have a great product (Dropbox) with an initial toehold in the European market that it wanted to expand. 

The landing team was a startup breaking ground in a new market, but it was also a flag-bearer for the Dropbox culture in a new office far away from San Francisco. 

“It was important for Dropbox to insure that the culture of the international offices was not exactly the same but shared the same core values as what we had at SF. A key element on how we chose the initial landing team was that they'd all be long-time Dropbox employees and excellent culture carriers for establishing that first office and onboarding its first hires.
Another conscious change we made was not using terms like HQ or satellite.  We were now a global company with offices in San Francisco and Dublin.  Reflecting back, I think subtle changes like this had a pretty big impact on evolving our culture and company to a global mindset."

Wang says he takes a project-based approach to growth. For international expansion especially, he looks for generalist problem-solvers who are willing to take on a variety of startup-style tasks. 

“One of the primary responsibilities for the bizops person on the landing team ended up being to do culture fit interviews. He would spend three to four hours a day on some days doing these interviews to ensure the team we built in Europe embodied the same core values as the employees in the U.S.” 

International might not be as different as you think

Japan was a big exception for Dropbox, but for other locations, the company started with a default U.S. model for distribution and growth. 

“Don't reinvent the playbook. What we found is that that actually carries you pretty far because it turns out a lot of international markets for Dropbox ended up being more similar than you might think: Norway, Sweden, Denmark, Finland, Australia, New Zealand, Canada and the UK were all fairly accessible to us using our existing approaches to sales and marketing.
The advantage with this first set of countries is that English will also get you ninety-five to one hundred percent of the way there, even if a totally localized version is still the ideal eventually.
As we moved to France, Germany, and Southern Europe, we started to see more specific regional and linguistic needs up front.”  

Or international could be really different. How to run growth experiments in a completely different “lab”

Experimentation is key to a productive growth function, and yet might not be as translatable as many U.S.-based teams would think.

How might the process of growth experimentation itself change when you’re dealing with new markets that might have different distribution platforms that are relevant to them, or different regulatory or social practices?

Wang says it depends on the sales life cycle of your product, starting from initial awareness and including how people transact and complete the sale. For example, in Japan, many people go to the equivalent of a Best Buy to purchase software. This is the purchasing norm that they're comfortable with, and it’s the distribution channel that Dropbox, Evernote and other companies — even though they’re online subscription companies — have adopted in order to cater to that market. 

“We partner with a local partner there who essentially makes a physical box and puts it on a shelf. Inside, there’s a card and a code that tells you to go online to redeem your Dropbox subscription. 
That's an example where the awareness is not online — it's in store. The transaction happens in store, not online. Even the payment system is through store, to partner, back to us as opposed to a direct credit card relationship.” 

The setup creates challenges for a recurring subscription product — how do you encourage or automate renewal? — and yet Wang insists that examining and challenging each step in the series is how you test for improvements in geo environments that look very different from the experiment environment you’re used to.

“You can take a related product to yours, the one that’s sort of a local equivalent to your product, and ask how how people purchase that product. Follow their flow step by step and you will glean a lot of information: how people learn about this product, how people prefer to make their purchases, where and so forth.” 

Get more insights on running growth experiments by email.

Data risks and regulatory norms

Another area often overlooked by U.S.-based companies going international has to do with data storage and ownership. For many business customers, there is a strong preference for data to be located outside the US. According to Wang, this is especially true in Europe, where news related the Patriot Act, NSA, Snowden and leaks has made U.S. companies seem like riskier keepers for data. 

“Germany in particular had sensitivities about this. Germany is a phenomenal market for Dropbox because of how many businesses are there, and how many of those businesses are collaborating with other businesses in other countries. However, data privacy with a U.S. company tends to be quite an issue for them.”

It’s related to being perceived as a local company versus a foreign one, says Wang. A lot of the hesitation ends up being about data privacy, not just customer support or sales norms, and much of it has less to do with specific regulatory and technical compliance than with trust.

For example, a German company may be sharing files with a design firm in Argentina, which means that the data will be shared into Argentina at some point. 

“In that case, it’s like ‘Oh ok yeah that's fine we just want to make sure our data is secure. It wasn't a technical requirement for how the data had to be stored in a certain way or not. A lot of it was the comfort they got from knowing that we were being thoughtful about data ownership and privacy as a local concern.” 

Currency and payments challenges with international expansion

Another area that the Dropbox team spent time optimizing was payments. Wang says his team was fortunate to learn from their peers at Evernote and SurveyMonkey, whose payments team offered valuable guidance in Dropbox’s early days of international. 

“With payments you have a few things, one is local currency. This goes back to perception once again. 
It makes a pretty big difference from a perception standpoint when you can charge customers in the local currency versus in US dollars.
Imagine if you’re paying for some monthly service and the amount on your credit card bill changes every month and you’re paying a foreign exchange fee on top of your product fee. That doesn't feel great. One of the first priorities when we expanded internationally was to offer local currency payments for Dropbox.”

While it feels more local and friendly, offering payment in different currencies poses unique challenges for forecasting growth, says Wang. 

“The Euro depreciating, Brexit, you name it so… all of this has brought up some interesting challenges from a forecasting standpoint. You need to make sure you have a good relationship with your finance team, and make sure that your operating team is not going to be either rewarded or penalized for international currency fluctuations. Let the finance team manage that, and you manage the communication with investors and the board and so forth. That has certainly been a roller coaster over the last few years.” 

Currency is one thing, says Wang, but payment methods is another beast. In Japan, Dropbox had to identify and rely on partners because the typical Western credit card purchase wasn’t going to work for their customers in that market. 

In France, Dropbox needed to add support for Carte Bleue (now Carte Bancaire), which functioned like a debit card without pre-authorization from the cardholder’s bank. 

In Germany, the credit card penetration rate is low (estimates put it at between 20-33%), so Dropbox offered direct debit connecting to the customer’s bank in order to withdraw money.

Since those earlier days, says Wang, there have been more payment startups to help ease parts of the payments process (Braintree, Adyen, Stripe), but most country launches still require a substantial amount of engineering investment to execute well.

Biggest mistakes in international launches

Dropbox is known for its data- and analysis-driven culture, which Wang says helped it to avoid some of the biggest mistakes in launching an international expansion effort. But a lot of companies view themselves that way, and yet many still end up falling prey to a set of common mistakes when going international.

The most foundational of these mistake, says Wang, is jumping into international expansion without enough research and planning. How much is enough?

“At Dropbox we did tons of planning — maybe we even overdid it and could have gone earlier — but we talked to other companies, asking them these same questions about mistakes, and one that stood out was doing it too fast.
For example, let’s say you made it your 2017 plan to ‘go international’ and your goal is to be in Europe in Q1, Japan in Q2, South American in Q3 and Australia in Q4. That's a mistake.”

Wang says it’s simply too much at the same time because of the demands on overhead, executive bandwidth and internal company mindshare that a multi-region international effort necessitates.

“It's just like building a product. You have to go through one cycle to learn and iterate and then the next cycle to learn and iterate. Don't try to do five simultaneous experiments at the same time. If you try to launch five products at the same time you're going to run into the same issue as launching in multiple countries at the same time.”

Some of the things that might come up, says Wang, include setbacks that don’t abide by your internal company timeline. For example, immigration visas can take longer than you expect, and at that point you’re waiting on governments, not merely a different team in your organization. Additionally, the various business licenses, commercial licenses, and other requirements for legal entity setup are all critical steps that you may not be able to accelerate or even influence.

Wang says it’s beneficial to have a dedicated "beachhead" office to help you liaise with the immigration office of the country you’re launching in. One benefit of Dublin, he says, is its transparency. It’s also a hub for international business, and is well-resourced with experienced immigration and legal firms that know how to move into other countries of Europe. Because of these infrastructure advantages, Dropbox’s initial step in Ireland helped them to succeed later in France, Germany, and the UK — but all of that would have been hard as a single mega first step.

In particular, says Wang, rushing to troubleshoot multiple external issues at the same time can create overload for your internal resources. 

“It’s not necessarily that overwhelming for the people on my team, business operations, or on sales. It's actually particularly overwhelming for ‘invisible’ functions at your company: the legal team, the finance team, and the HR team. 
I remember there were a lot of conversations about what was the right ex-pat package. How do you translate healthcare benefits over for ex-pats versus for local hires? A lot of these things are complicated but important questions, and solving them creates a lot of extra work for the people you don't initially think are required to be part of an international effort.” 

These are all reasons for why anyone tasked with international expansion should have strong cross-functional skills, and good social capital within the company, says Wang. You will be imposing a lot of extra work onto everyone else, and the stronger your relationships the better the outcomes. One way to burn those relationships, asserts Wang, is to aim for too much, too fast.

“Get one round out there. Internalize the lessons and build your V1 international playbook, then apply the playbook to make the second round faster, the third round even faster and go from there. But don't go from zero to ten, because it's going to be a big struggle.”

Special thanks to Chenli Wang for his participation in this post.