This is the fourth in a series on building a global company from the ground up.

A startup within a startup

Doing a startup is really hard.  No one in their right mind would consider doing two startups in parallel, would they?  And yet that’s what going international is:  a second startup inside your first one.  And like any startup, it needs people and money to succeed.

Two key questions

Once you’ve got your strategy (see Part 2 for more on strategy), you still need to answer two key questions before you can pull the trigger on international expansion:

  1. How are we going to staff it?  There are two pieces to this—Visible and Invisible:
    • Visible: new people dedicated to international, whether at HQ or in country.
    • Invisible, but just as critical: existing people in the current organization who will need to do additional work to enable it—engineers for internationalization, support staff for ticket escalation, HR people for recruiting and comp work, lawyers for incorporation and tax strategy, etc.
  2. How are we going to fund it?  Again, two pieces:
    • Visible: direct costs for new people, facilities, travel, etc.
    • Invisible: the hidden costs of the additional work required from the rest of the organization.

While you can’t ignore the visible direct costs, there’s a great temptation to ignore the invisible piece: “Come on, we have a great team, they’ll just get it done.”  That’s a big mistake.

In my first international startup role, I experienced that first-hand.

@Home – From beggar to banker

In early 1997, I was recruited to start up the international business for @Home Network.  It was a unique opportunity.  The two-year-old domestic business was growing like a weed, serving American consumers’ insatiable desire for the broadband Internet over cable the company had invented.  Investor interest was equally high, and within three months of my joining, ATHM had become the newest high-flying stock on the NASDAQ, with a market cap of $30 billion on $48 million of projected 1998 revenue (these were, after all, the dot-com days).  The domestic business was on a defined track to profitability, with a plan that projected steadily shrinking quarterly losses and a near-term crossover into the black.  Adding in the losses from my international “startup within a startup” would reverse that trend and push out profitability by several years.  And therein lay my problem as the founder and so far only employee of @Home’s international business—a problem common to any international startup inside a domestic parent, whether private or public.

The more new countries we launched, the greater the pain would be—and while Wall Street loved the international growth story, they didn’t want to hear anything about more losses for a stock already priced to perfection.  What’s more, the organization was already stretched to keep up with the intense demands of the exploding U.S. market.  As a result, I enjoyed high expectations and lots of goodwill, but no resources to internationalize code and content, deploy network infrastructure, hire and train overseas teams, or anything else.  Once we had an international strategy, we had to find a way to staff and fund it.  Here’s how we did it.

For each country to be entered, we first developed a detailed two-year startup budget.   Naturally, the budget included the dedicated headcount and other spending we would need to operate in country, but we also budgeted additional heads in each of the HQ functions whose support we’d need to get up and running.  For example, if we needed three months of engineering work to support Japanese cable modems, we explicitly included a quarter of a headcount for that in the budget.  If we expected the HQ support organization to provide Level 3 and 4 escalation support for our Australian operation, we budgeted heads for that too.  If we needed a facilities guy to fly to the Netherlands to spec out a data center, we included that too.  Having quantified the needs, the next challenge was how to fund them.

We found a solution by tapping into the appetite of overseas cable operators, whose cooperation we needed in any case to deliver our broadband Internet service.  I generally don’t recommend joint ventures, but in this case they made sense.  For each country, we formed a local venture (e.g. @Home Japan) jointly owned by @Home and the local cable partner.  The local venture then awarded the @Home parent company a paid 18-month consulting contract to manage the startup—from finding an office to designing the network and server infrastructure to recruiting and training the local team.  The local partner provided the funding (sometimes with some capital contribution from @Home).  Although they grumbled about having to fund @Home’s startup resources, they ultimately realized that this was the way to get up and running faster and with higher quality than any other approach.

This approach transformed my new international organization from beggar to banker.  Instead of going cap-in-hand to beg for support from the VP of engineering, we could tell him, “I can fund two new heads for you, and all their T&E for the trips they’ll make for the U.K. startup.  As long as you meet our service level agreement, it’s up to you what you do with those two extra heads.”  In many cases, VPs would choose to assign their most experienced high-performers temporarily to oversee the startup (which the employees viewed as a nice perk) and used the funding to backfill them with new hires.  In a headcount-constrained company, this approach made international projects very popular instead of an unfunded headache.  It also enabled us to get multiple ventures up and running and transitioned over to local teams quickly and effectively, which made our partners and our investors happy too.  The consulting dollars even gave us international revenues right away, to be replaced by royalties and other direct revenues once the business took off.   Within a couple of years, we had over a quarter of a million broadband subscribers outside the United States.

In short, at @Home we figured out a way to cover the invisible people and costs required to make international successful.  On the other hand, when @Home acquired Excite for $5.7 billion in 1999 and I added responsibility for their international business, I got to see the painful consequences of ignoring the invisible.

Excite – Ignoring the invisible

Like @Home, Excite (a major web portal) was already public and had launched in Europe and Japan by the time we acquired them.  Unlike our international business, however, Excite’s seemed to be pretty unpopular, both inside and outside the company.  The general manager, whom I’ll call Dick, seemed highly stressed when I met him, and for good reason, as it turned out.  Excite had hired 20 to 30 staff in five countries and formed JVs with British Telecom, Telecom Italia and Itochu—and they weren’t getting the support they’d been led to expect from Excite corporate.  While the JV partners were funding some of the local operations, there was no budget or headcount plan for Excite’s support of the international business.  Dick was literally the only California employee with any responsibility for meeting the needs of five country subsidiaries and three large, demanding partners.  Initial goodwill on the part of other departments towards the overseas business soon turned to negativity, as the list of internationalization requests grew longer and it became clear that there was no extra headcount or budget to support them.  The result was weak local ventures, frustrated partners, low morale and a cloud over what should have been an exciting expansion of the business.  As for poor Dick, he was practically afraid to answer his phone or read his email anymore, such was the mismatch between the expectations created and his ability to deliver.

The Excite case is not some exceptional example of particularly bad judgment or poor execution—just an illustration of what is almost guaranteed to happen if you don’t think through your international expansion all the way.

Several ways to proceed

There are ways to replicate the successful @Home approach above without doing joint ventures.  At Opsware, we negotiated a distribution deal with a large Japanese partner, with $5 million in minimum commitments that allowed us to fund the product work, training and local hiring required to enter the Japanese market.  At Silver Spring Networks, on the other hand, we bit the bullet and funded it directly.  In every case, however, we mapped out up-front all of the work required to make our international startups successful and assigned a funded budget for it—down to the finance, legal and HR heads we would need to draw on for tax, incorporation and compensation advice.

Remember—if it’s unfunded, it’s not going to get done.

Up next: Wrap-up

Comments or questions?

I welcome your comments and questions.  If the comments box is not showing up below, click on the title at the top of the post and it should appear.

This is the third in a series on building a global company from the ground up.

In my last post, I discussed the importance of developing an explicit international strategy.  But who should drive the process, and how can you best achieve commitment to its execution?

The Loudcloud experience

The first couple of years of Loudcloud provide a useful illustration of the challenges of rapid international expansion, even in the best-managed companies.  The business had taken off like a rocketship from its founding in late 1999—revenue grew from  $2M to $55M in a year.  While that growth was exciting, it led to a very high financing valuation for such a young company: $820M for a company just nine months old.  High valuations mean high expectations, and bookings expectations for Loudcloud were sky-high.  To add to our challenges, we had over 30 competitors rushing to grab market share.  Faced with a large bookings target and nearly unlimited funding, our head of sales came up with a plan to expand internationally, opening European offices and hiring aggressively to build the business as rapidly as possible.  The entire business was moving at breakneck speed, and the other functions of the company (including sales management) had their hands full just meeting US customer demands.  As a result, the European offices didn’t all get the time and support they needed to achieve productivity.

In the extraordinary circumstances of the time, that was understandable, but it proved costly when the rules changed.  When the recession of 2001 hit, we had three fairly expensive European sales offices with varying levels of productivity and maturity.  While the UK office was contributing, our offices in France and Germany were burning cash that we didn’t have, with few bookings to show for it.  We ended up shutting them down, at a cost of major management distraction and substantial severance and legal expenses, especially related to exiting France.

In any company, and especially in a startup, there are far more things to be done than time or resources to do them.  In a high-growth company, the core domestic business will inevitably tend to consume most of management’s attention and the support of the company’s functions.  Taking the company international is a long-term project that demands sustained effort.  Your international aspirations will go unfulfilled unless you start with an explicit strategy, supported by all of the functions of the company; adequately resourced; and led by an executive with the ability, authority and time to pursue the international business.

Assign an owner from the beginning

Down the road a year or two, once the international business is established, you may decide to appoint a worldwide VP of Sales or Field Operations to run it, or in some cases create a separate International division.  At Opsware, for example, once we hired Mark Cranney (now my partner running Market Development at Andreessen Horowitz) to run sales, we also gave him responsibility for international operations.  However, in a startup situation, I recommend giving the initial responsibility to someone else on your team if possible – particularly if the core business is still heavily focused on the U.S. market. You don’t want quarterly bookings to suffer, or your fledgling global strategy to get hurt by short-term sales pressures, so you should probably not give the role to your VP of Sales, even if she really wants it.

This person’s role is to:

  • Be the voice of international on your team and throughout the company.
  • Become the company’s expert on overseas trends, competitors, partners, users, customers.
  • Get out of HQ: From a very early stage in the company’s life, make frequent visits to key overseas markets to gain first-hand knowledge and build the company’s reputation and relationships – and set up productive, efficient visits for you and other executive team members.
  • Represent the company at overseas industry events.
  • Lead the development of the international strategy, leveraging all of the knowledge gained, and involving all key groups in the process.
  • Lead discussions and negotiations with prospective partners, bringing you and others in as appropriate.
  • Recruit and lead a small global startup team.
  • Potentially manage the initial startup in the first several countries.

Ideally, you want someone with international experience, who knows the company well and is seen as passionate, credible and impartial by the rest of the team. They should be someone you trust to speak at overseas conferences and to represent you in negotiations with partners or launch customers around the world.

This is one of the roles I played for @Home and for Silver Spring Networks, and to some extent for Opsware in its early years. The role demands a ton of time and frequent international travel—at Silver Spring, I estimate my team and I spent 30-40% of our time on the road, with the other 60-70% evangelizing and building capability back at HQ.

Given the time demands, this person shouldn’t have day-to-day operating responsibilities that could suffer.  In my experience, this is a natural role for your VP of Business Development, if you have one that meets these criteria.  (If you don’t, you may need to find one!)

Involvement builds commitment

Your international strategy should be the roadmap that guides the whole company’s pursuit and support of the global opportunity over the next 3-5 years.  Like any strategy, it will be useful if it’s relevant and realistic, which it will be if all key groups in the company have a hand in developing it—from Sales, Product Management and Engineering to HR, Legal and Finance.  Your BD VP’s role is to provide facts and data and context, and to drive a process that enables their participation in an efficient manner.

Ultimately, you own the company’s strategy as the CEO.  Make it clear that becoming global is a critical priority for you, and participate actively in the process.  Broad involvement means broad commitment to the result.  Don’t let it be an academic “staff exercise”!

At Silver Spring Networks, developing a version of our advanced wireless technology to meet European spectrum requirements was a critical element of our strategy to capture the emerging international smart grid market.  We involved key engineers directly in understanding the business opportunity—not just the technical requirements—and exploring strategic alternatives.  Thanks to this involvement, the engineering team became key partners in enabling our international success, despite a busy roadmap for the U.S. business.

From strategy to action

Once the strategy work is done, roll it out to the company and make it your personal mission as CEO to make it sure it’s translated into action.

I’ll talk about that in my next post.

Up next: Execution

Comments or questions?

I welcome your comments and questions.  If the comments box is not showing up below, click on the title at the top of the post and it should appear.

Twenty-five years ago, I funded my Stanford MBA by becoming a junkie—a spreadsheet junkie, that is. IBM had just acquired a high-flying Silicon Valley technology company called ROLM, and the IBM execs needed data and analysis to manage and track their progress against their business goals. As a young engineer just entering the business world, I was surprised by just how extraordinarily difficult it was for managers to obtain relevant information on a timely basis to guide their decisions—even in the corporation that invented business computing.

Their difficulty was my opportunity, and I earned many thousands of dollars writing ever more sophisticated Lotus 1-2-3 macros to automate the extraction and analysis of market data for my new white-shirted bosses from Armonk, NY. The money’s long gone, but the key lesson from that time remained with me as I advanced into more senior management roles in six subsequent companies: for a manager trying to make high-quality fact-based decisions every day, it’s really hard to get good data.

Business Intelligence – Big Market, Low Satisfaction

It turns out that it’s worth a lot of money to customers to try to solve this problem. In fact, their willingness to spend has funded a lot more than just my MBA in the last 25 years. It’s spawned a $25 billion industry known as Business Intelligence, or BI, devoted to extracting and making sense of the masses of data trapped in corporate IT systems, with a couple of billion dollar revenue companies (Business Objects and Cognos) and one half-billion dollar company (Hyperion), along with thousands of smaller software companies and consulting firms.

But these companies achieved this level of success without even really solving the problem. For many IT buyers, traditional BI solutions came to epitomize everything that was wrong with large-scale enterprise software.

  • Just to get started required a spend of at least a million dollars on software and hardware, with maybe another million on consultants for deployment and integration with a separate data warehouse, as well as the unwieldy ERP or accounting systems that were the source of the data.
  • Maintenance and upgrades cost another 20% a year.
  • Deployment would take six to 12 months, by which time user requirements might have changed.
  • Trying to add a field or two to a dashboard, or create that new report format requested by the CEO might take weeks because changes and ongoing maintenance typically depended on a small team of dedicated experts with deep knowledge of the back-office systems and BI solution.

It’s no wonder, then, that many traditional BI implementations fell well short of their promise—confined to a small set of users and dashboards, or even shelved entirely.

As a result, that first wave of big BI pioneers is no longer with us, at least not as independent companies (SAP bought Business Objects for $6.8B, Oracle bought Hyperion for $4.5B, and IBM bought Cognos for $4.9B—all in 2007). In their defense, they did a pretty remarkable job with the technology and software business models of the time and created billions of dollars of value for their shareholders. Furthermore, their solutions continue to generate sizeable legacy software sales and consulting revenues for their new parents.

BI in the Cloud – The Chance to Do it Right

Business users’ appetite for timely, relevant information and analysis for decision-making hasn’t diminished since I wrote my first Lotus macro back in 1984. In fact, it’s stronger than ever. What has changed fundamentally in the last few years is our ability to meet enterprise customer needs efficiently and effectively.

The SaaS delivery model means the customer starts to see value from their software within days or weeks, with an up-front investment of thousands, not millions of dollars. Changes and upgrades are seamless, and included in the monthly subscription. The cloud enables consumption to be scaled up and down on demand on a pay-as-you-go basis. The platform shift to SaaS and cloud computing creates an opportunity to finally deliver on the promise of Business Intelligence.

Like their peers in other large established software categories, the old BI vendors are valiantly striving to port their legacy client-server stacks to the cloud. In any platform shift, however, the big winners are the new entrants who designed their product and business model around the new platform from the beginning.

Introducing GoodData

Just like Salesforce.com in sales-force automation and Netsuite in cloud-based ERP, we believe there’s an emerging opportunity for the right new entrant to build a giant new franchise in BI. We’re putting our money on GoodData. We’ve followed the company closely since we invested in their original seed round, and we’re delighted now to be leading a $15 million expansion round to capitalize on the strong lead they’ve established in BI for the cloud.

Our enthusiasm starts, as always, with the Entrepreneur. CEO Roman Stanek is that powerful combination of brilliant product visionary and compelling sales guy. He started two successful software companies, Netbeans and Systinet, from his native Czech Republic prior to founding GoodData in 2007. He’s assembled a stellar team to pursue the BI opportunity and they’re executing very well.

We’ve been highly impressed with GoodData’s Product. Like its market-leading peers in other SaaS categories, GoodData was built on a multi-tenant platform designed expressly for the cloud. Unlike other BI offerings, it’s a complete solution, avoiding the traditional struggle to integrate one vendor’s BI product with another vendor’s data warehouse. While it can be implemented within days, the product is built for large scale. That’s critically important, because when you offer business users true operational business intelligence on demand, adoption and utilization grow explosively. (For example, there were 2 million report executions on the GoodData platform in July, up almost tenfold in six months.)

GoodData has already signed more than 100 direct and 2,500 indirect Customers, including mid-sized enterprises like Enterasys Networks and Pandora, as well as large corporations like Time Warner Cable and Capgemini. When we spoke to them, they said things like:

“The great thing about GoodData is that it’s been valuable right from the beginning – no waiting game…Our sales team now has direct access to metrics that used to take days for our engineering team to produce.”

“With GoodData we don’t have to jump through hoops to view our critical data.”

“GoodData has rapidly become our source of truth.”

“It’s put the power in the hands of our end-users.”

GoodData’s target is the ever-increasing number of established enterprise customers now migrating to cloud apps like Salesforce, Netsuite, Google Apps and Zendesk, as well as the new generation of players like Pandora who have built their entire business on the cloud from the beginning. Because cloud applications are dramatically easier and cheaper to implement, customers are deploying far more software than they did in the old world—one customer we talked to already has 18 cloud apps.

Roman and his team realized from the outset that it all starts with Apps: a cloud-based BI solution that provided dashboards, reporting and analytics on top of individual cloud apps as well as across apps—for example, combining website data from Google Analytics, marketing analytics from Marketo and CRM data from Salesforce to produce an integrated Lead to Cash report—would be hugely valuable to these customers. They already support most of the leading SaaS and cloud apps, and the number is increasing by the week.

However, they’ve gone further than that, creating an innovative Partner program called Powered by GoodData that allows leading SaaS and cloud providers to embed dashboards and analytics directly into their own applications. For example, help-desk software leader Zendesk—itself one of the most impressive exemplars of the new generation of cloud-based app providers—makes a compelling solution even more compelling by providing GoodData’s advanced dashboarding and analytics to its best customers.

In Conclusion

Business Intelligence, and IT as a whole, has come a long way since my days as a spreadsheet junkie—and the best is yet to come as the shift to the cloud becomes mainstream. Management’s demand for relevant and timely information for business decision-making will continue to grow, and cloud-based BI will make it accessible to more users, in more companies, than would ever have been possible under the old regime.

As veterans of the original cloud computing company, Loudcloud, we’re excited to add GoodData to the list of Andreessen Horowitz-backed companies that are leading the charge to that brave new world.

This is the second in a series on building a global company from the ground up.

In the first part of this series I discussed the importance of laying the groundwork for a global company up-front.  In this part, I’ll talk about developing an explicit strategy to guide your international expansion.

What’s your international strategy? 

International expansion can happen almost by accident—if you’re a consumer Internet company, you get offered a deal by a big German partner, or if you’re an enterprise software company, your sales guys sign a bluebird deal in Japan to make the quarter—but that rarely turns out well.  The markets you end up in turn out to have limited potential, or the cost to serve them turns out to be a lot greater than you thought.  Customer commitments get missed, or tensions arise between the sales guys who sold the deal and the product people who weren’t involved but now have to deliver it.  All of a sudden, international expansion has a bad name.

It doesn’t have to be like that.  On the contrary, going global can be highly rewarding  for everyone if it’s deliberate and strategic.  Before you make your first move overseas, you should have an explicit, well-communicated, company-wide international strategy that will act as the company’s “True North” as it crosses the border.

Five key questions

Your strategy should address:

  1. What’s our goal?
  2. What countries will we focus on?
  3. With what product(s)?
  4. How will we go to market?
  5. What’s our operating model?

Let’s talk about each in turn:

1.  What’s our goal?  This can be a qualitative statement such as “Be the number 1 player in mobile gaming in the top 5 world markets by 2014”, or a quantitative target such as “Generate 50% of our revenue from outside the US by 2016”—or a combination of both.  It should be big, specific and measurable, because it will set the bar for the rest of the strategy.

2.  What countries?  What countries to focus on, and in what order, is perhaps your most important decision.  Many US companies launch in the UK next, because they speak English and it feels familiar – but is that the next best market for you?  Every new country you enter will require an investment of precious time and money, and will generate a certain return.  In simple terms, you want to prioritize countries offering the most return for the least investment.

That should start with a quantitative, data-driven analysis (aka a spreadsheet!), ranking the world’s countries according to the criteria that define an attractive market for you.

Those should include both Macro criteria related to overall market size—such as Population; GDP per Capita; Internet penetration; Broadband penetration; Smartphone penetration; 3G deployment1—as well as Company-specific criteria.  If you’re an enterprise software company targeting financial services giants, then “Number of banks with over $XB in assets” would be a highly relevant criterion.  For our data center automation company, Opsware, it was Number of servers or Network devices.  Sometimes fairly arcane factors can make one market much more attractive than another:  When I was running @Home International, Availability of two-way cable networks for broadband Internet services made markets like Australia and Netherlands more attractive than much larger countries like Germany.  Competition will almost certainly be a criterion, as will Product fit:  Your product may work with minimal changes in some countries, while requiring months of development work for others. 

This is real work, but it’s a hugely helpful exercise, because it forces everyone involved to think through what makes an attractive market, in terms of both potential and the investment required to realize it.  When you’re done, you should have a ranked top 20 list that will both guide your proactive expansion and give you a fact-based framework to react to opportunities that pop up along the way.

Naturally, analysis doesn’t take the place of judgment—it just supplements it.  You may decide to postpone entry into a market that looks good on the spreadsheet because you believe the company isn’t ready or the risk is too high.  For example, China might look very attractive on the numbers, but concerns about IP protection or ability to enforce contracts might well deter you.

3.  What products?  Obviously, you need product requirements and a product plan for every market you’re going to enter.  Unless you’re lucky enough to have a product like Twitter that works pretty much the same everywhere, that means product management and engineering work, and additional resources to be assigned to getting it done.

4.  How will we go to market in each of the top 3-5 countries?  Another question with major ramifications.  If you’re a mobile app company, maybe a partnership with a mobile operator is the way to go.  If an enterprise company, will you need to hire a direct sales force or can you start with a channel partner?  If you have a freemium model like Box.Net or Splunk, maybe you can seed the top 10 countries by promoting free downloads and then target the most successful ones with direct sales.

If local clones have got there before you (as in Groupon’s case) you have a build/buy decision to make:  Acquiring a local competitor buys you instant market position and a team, but technology that may be hard to integrate and a team you didn’t hire, with values that you may not share.

While every situation is different, and there’s no one right answer, I would generally beware of relying excessively on a partner, and particularly try to avoid joint ventures—more on that another time!

5.  What’s our operating model?  In the rush to launch overseas, this is often the most overlooked question.  The core decision you need to make is which functions you’re going to manage centrally, and which locally.  For example, if a shared global technology platform is a fundamental competitive advantage (as it is for Google, Facebook, Zynga, or Airbnb), then that’s one area where you don’t want to give local autonomy.  In its early years, Yahoo! gave its individual European country teams virtually free rein to develop local versions, leading to a wasteful hodge-podge of inconsistent platforms, petty fiefdoms and internal infighting between European and corporate executives.

Source:  Campaign UK, February 19, 2001

While Google powered ahead with a common technology platform and a clear command structure, Yahoo! Europe never recovered from this failure to pick the right operating model.  Read between the lines of this extract from Carol Bartz’s February 2009 reorg memo:

Regions: There are now two: North America and International. As I’ve said before, international growth is critical for Yahoo!, which has become too reliant on its U.S. business over the years.

The regions deliver Yahoo!’s products, programming and services to consumers, partners and advertisers in local markets. They will partner closely with the newly formed Regional Solutions & Products group in Ari’s organization to help drive a significant shift in how Yahoo! develops products for different geographies. The goal is to have global platforms on which regional product offerings are based.

The North American region—comprised of the U.S. and Canada—is led by Hilary Schneider. The leader of our International region, to be hired soon, will be responsible for a cohesive Yahoo! global strategy and seizing our international growth opportunities. Until we determine who’ll lead the International region, Rose Tsou (Asia), Rich Riley (Europe) and Keith Nilsson (Emerging Markets) will continue to report to me.

Get the sense that an explicit international strategy up-front might have helped?

Operating model decisions drive organization structure decisions:  In most cases, a matrixed structure will ultimately provide the best combination of centralized consistency and efficiency with localized customization and execution:  Local team members reporting dotted-line to a local GM and hard-line to their corporate function (or vice-versa). The org structure will need to change over time as the international business grows in size and complexity.  At the beginning, you should optimize for low overhead and rapid execution, but with enough corporate legal and finance oversight to minimize the risk of bad tax or IP decisions, or worse, fraud or FCPA violations.

Wait—that’s a lot of work!

By now you’re undoubtedly thinking:  I’m already working 16 hours a day, and so is my team—who’s going to do all of this new work?  As the CEO, your role is to sponsor the work and assign someone passionate, credible and independent to lead it.  I’ll talk about that in my next post.

Up next: Assigning ownership and ensuring commitment.

Comments or questions?

I welcome your comments and questions!  If the comments box is not showing up below, click on the title of the post and it should appear.


1 There are lots of good free sources for this data, e.g. CIA World Factbook; ITU; OECD, as well as dedicated sites like Internet World Stats.

This is the first in a series on building a global company from the ground up.

In browsing Quora recently, I was struck, but not surprised, by how many people sought advice on how to expand a startup internationally.  Not surprised, because for many young companies international expansion ranks with M&A as something that’s easy in concept, but difficult to execute well.  (See the many Quora responses, including my own, to Why do companies often fail at international expansion? )

I believe you can maximize your chances of getting international right the first time around if you:

  1. Lay the groundwork from the moment you start your company
  2. Launch internationally with an owner, a strategy and a plan
  3. Apply some key principles and best practices once you’ve launched

I’ll devote a blog post to each of these topics, starting today with the first one.

Part 1: Laying the groundwork

I grew up in Ireland.  With a population of only 4.5 million, it’s a tiny market—so Irish entrepreneurs have to think outside their borders from the beginning.  Relative to his Irish counterparts, the American entrepreneur is born with a silver spoon in his mouth.  He has the luxury of a massive home market—300 million affluent consumers, 30 million businesses, one language, one currency, one culture, one legal system—from sea to shining sea.  Initially, that’s a huge advantage that allows him to build a company of significant size without even needing a passport.  Google rocketed from zero to almost $350M in revenue in four years—80% of it from the United States market.

On the other hand, that huge starting advantage can become a liability when it comes time to expand internationally.  I’ve personally experienced the pain of taking products, systems and processes designed for the US market and trying to re-engineer them years later to work in Europe, Latin America or Asia.  Even Google struggled initially—more on that below.  The good news is that much of this pain can be avoided if you lay the groundwork from the time you start the company.

Talk the talk from the beginning—and walk it too

Even if you’re still far from ready to launch overseas, it’s never too early to start talking and acting as a future global leader.  Based on my experience, here are some ideas to consider:

  • Make the company’s mission explicitly global from the outset.  Mission statements like these set a global tone (the underlining is mine):

GoogleTo organize the world‘s information and make it universally accessible and useful.

Nike:  To bring inspiration and innovation to every athlete in the world.

AvayaProvide the world’s best communications solutions that enable businesses to excel.

FacebookTo give people the power to share and make the world more open and connected.

  • Mix in some international DNA on your management team—consider making international experience an explicit hiring criterion for at least, say, 30% of your positions.  In addition to international perspective, your team will benefit from diversity of thinking.
  • Make internationalization and localization experience mandatory for senior product management and engineering hires.
  • Recruit board members with international experience and perspective.  At my last company, Silver Spring Networks, we killed two birds with one stone, adding the CFO of Nokia to our board for both his international and financial expertise.
  • As you develop product and market strategy for the US market, challenge your team to think about the implications of taking this beyond the US later.  What might we do differently if we wanted to take this to Asia next year?
  • From the beginning, try to track relevant developments in key overseas markets:
    • What countries are we getting most website traffic from? Why?
    • Ask your employees who are citizens of other countries to keep you posted on what’s happening in their home country—they’ll be delighted to do it.
    • At least once a year, make an effort to spend a few days in Europe, or Brazil, or India or China. You’ll come back with a hundred new ideas.
      • Talk to local competitors, customers, partners, investors.
      • Is this market developing differently from or similarly to the home market?
      • What local players are emerging that we need to watch?
      • What user or customer trends offer lessons to improve our product or market strategy?
      • Who might be a great local partner to get started with?
      • Reinforce the global leadership message in all-hands meetings, company emails, one-on-ones, performance reviews—get the message out any way you can, and keep doing it!  Our market is the world!
      • Carve out some time to discuss international expansion at quarterly reviews or strategic offsites.

Make the right code choices from the start

Many entrepreneurs don’t think about it, but they’re making a decision with major ramifications for their global strategy from the moment they write their first line of code.  I was amazed recently when a senior executive from a major emerging Internet company told me that they had neglected to internationalize their code when they started, which would now cost them millions of dollars and months of delay in getting into overseas markets.  In the meantime, local competitors were springing up around the globe.  Don’t make that mistake.

Here are some suggestions:

  • Read the excellent “Beyond Borders” by John Yunker.   
  • Make an explicit decision on internationalization up-front.  Ideally, your code should be fully internationalized, meaning that it can be adapted to different languages and regions (aka localized) without engineering involvement.
  • If your code isn’t internationalized, the problem is getting worse with every line you add.  Bite the bullet and get it done as early as possible.  I highly recommend this Quora post by Yishan Wong about Facebook’s organization-wide effort in 2007 to separate their content from their code and set it up for crowdsourced translation into over a hundred languages.   
  • Facebook’s brilliant use of crowdsourced translation has become legendary among consumer Internet companies.  However, enterprise software companies are also making effective use of their global user community for translation, as Nicholas Muldoon of Atlassian Software explains in his Quora post.  Check out https://translations.atlassian.com.
  • Unless you’re a giant organization like Facebook, it won’t make sense to create your own translation platform.  While it may not work for everyone, I was intrigued when I got a chance to see Smartling’s approach to dynamic website translation. 
  • Although customer-facing code may be the most obvious internationalization candidate, your revenue-related systems are equally critical:  By 2004, more than half of Google’s traffic was international, and the company could boast of 104 language versions and 13 overseas offices—but monetizing it proved challenging initially, due to systems that were never intended for international use.  With a billing system described internally as “fragile” and “glued together”, “the company worried that it might not be able to accept payments from advertisers in some foreign currencies as well as through bank transfer and direct debit in certain locations.”  Don’t let that happen to you.

Protect your IP 

  • Make sure your IP protection plan covers you internationally.  (The Patent Cooperation Treaty process can give you some global protection while postponing the need to file regional or national patents.)
  • Register trademarks and domain names in the top 10-20 world economies.  You don’t want to invest in building a killer global brand only to find someone registered it before you in Russia.

What’s next?

If you take these steps, you should be in excellent shape to execute a winning global strategy when you’re ready to do it.  Of course, to execute a winning global strategy, you have to have a strategy and someone to lead its execution.

I’ll talk about that in my next post.

Up next: Assigning ownership and developing a global strategy.

Give a girl the correct footwear and she can conquer the world
—Bette Midler

I did not have three thousand pairs of shoes, I had one thousand and sixty
—Imelda Marcos

As entrepreneurs who experienced the first wave of the Internet, we regularly recognize ideas from the creative ferment of the late ‘90s being pitched to us in a new form by today’s entrepreneurs. The newest member of the Andreessen Horowitz family, however, is reinventing a concept pioneered by a 1920s startup—and doing so with dazzling success.  (Read yesterday’s post about old ideas made new again.)

A Jazz Age Startup

Back in 1926, an economist and advertising man named Harry Scherman and two co-founders established the Book of the Month Club.  Convinced that bookstores were not catering well to a rapidly-growing and geographically dispersed American middle class with an insatiable desire for literature, these 1920s entrepreneurs offered readers a revolutionary mail-order subscription book service.  Subscribers agreed to purchase a book recommended by the club every month, with no title costing more than three dollars.  If a customer didn’t like that month’s selection, they could swap it for an alternate title from the club’s list.

Scherman pioneered several remarkable innovations that contributed dramatically to the club’s remarkable success.  First, he built the club’s brand as a definitive arbiter of quality and taste.  The “Book of the Month” was chosen by an expert Selecting Committee made up of five famous writers and critics—the literary celebrities of the day.  Being published in a club-branded edition as a “Book of the Month Club Selection” was enough to propel a previously unknown author to mass popularity.  Second, he applied the principle of  “negative response”, in which subscribers would automatically receive a book every month unless they proactively opted out.  Finally, he tapped into consumers’ love of serendipity:  Club members would eagerly anticipate the arrival of each month’s new selection, chosen for them by the “curators” of the Selection Committee.

Scherman’s subscription book business brought millions of affordable, quality titles to a subscriber base of over half a million households by mid-century, and spawned hundreds of mail-order subscription successors throughout the twentieth-century, from Oprah’s Book Club to clubs for CDs and DVDs.   In a sign of a truly great idea, however, Scherman’s concepts are also the basis for an amazing new twenty-first century ecommerce business.

From Jazz to Sole—Introducing ShoeDazzle

I’m proud that Andreessen Horowitz is leading a $40 million growth investment in a company that’s leveraging the Internet to reinvent the fashion business in the same way that Scherman used the postal system to reinvent the book business.  The company is ShoeDazzle, and, while the business is women’s shoes and accessories and many of the concepts are new, the parallels to the Book of the Month Club and its brilliant execution are remarkable.

Every new ShoeDazzle member (now more than 3 million strong!) first takes an engaging 3-minute “style quiz” that allows the company’s stylists to create a monthly selection of shoes and accessories personalized to her individual style profile.  She then receives an email invitation to a personal online “showroom” based on her style profile, with a new selection on the first day of every month.  She chooses her favorite item (or items) for a standard $39.95, and receives them in a beautiful package within days.   She can request an alternate selection, or skip one or more months if she chooses, so the service is highly flexible.

Like Scherman with bookstores, ShoeDazzle’s founder and CEO Brian Lee realized that traditional e-commerce sites offering women’s fashion were missing the point.

Observing the pleasure his wife took from visiting a favorite boutique to buy her latest pair of shoes, Brian and his team set out to recreate and even improve upon that deeply satisfying shopping experience on the Web:

  • They recognized that many women go shopping because they enjoy it, so they focused on making the experience fun and entertaining rather than purely functional.
  • They used the concept of personal stylists to replicate the trust that women feel when they visit their favorite store.
  • Like Scherman with his committee of famous authors and critics, Brian recognized the influence of celebrities on women’s fashion choices, enlisting Kim Kardashian as his co-founder and recruiting an ever-growing roster of additional celebrities as stylists and sponsors.
  • Like their 1920s counterparts, Brian and his team realized the appeal of serendipity and anticipation.  ShoeDazzle members await the first of the month with its new selection with the same avid excitement Book of the Month Club members must have felt as they awaited the mailman with their latest monthly title.  Unlike their Jazz Age predecessors, they can share their experience of opening the box with the whole world.
  • Whereas the closest those 1920s entrepreneurs got to social networks was probably local book–reading clubs, the ShoeDazzle team has been able to leverage Facebook in a truly impressive way to create a truly remarkable social shopping experience and an extraordinary community of almost a million enthusiastic fans.
  • Like Scherman, Brian and team understood the importance of brand, and brand all their products with a ShoeDazzle name that stands for exciting, affordable, quality fashion products, curated by experts and backed by extraordinary customer service.

Transforming the way fashion products are marketed and sold

From Valpak to Groupon, from the Full Service Network to the Internet, from Excite to Google, from the Book of The Month Club to ShoeDazzle—the story of the technology business is often a story of old ideas made new again.  Like their illustrious predecessors, Brian Lee and his team have taken a great old idea, added some brilliant new ideas of their own, put it all into Twain’s “mental kaleidoscope”, and come up with a “new and curious combination” they call ShoeDazzle.  The result is a new and genuinely exciting approach to ecommerce that is transforming the way fashion products are marketed and sold.

We’re proud to be associated with the ShoeDazzle team, and we can’t wait to see where they take it from here.

“There is no such thing as a new idea. It is impossible. We simply take a lot of old ideas and put them into a sort of mental kaleidoscope. We give them a turn and they make new and curious combinations.”
—Mark Twain

Although Mark Twain’s statement that “There is no such thing as a new idea” may be an exaggeration, some of the world’s best businesses have been built on leveraging old ideas in a “new and curious combination”.  Some great ideas work spectacularly the first time around, handsomely rewarding the original entrepreneurs.  Others fail or flounder initially, sometimes multiple times, before a combination of the right entrepreneur and the right market and technology conditions unlocks their true potential.  Some of the best ideas of all have a great first run, only to return decades later and succeed all over again—reinvigorated with the latest technology and fresh thinking by a new generation of entrepreneurs who may not even be aware they’re leveraging an old idea.

One of the things I love about the technology business is seeing old ideas made new again.   It worked before—will it work again?  It didn’t work before—will it work this time?  At Andreessen Horowitz, it’s not unusual to see ideas from the first Internet wave of the late ‘90s.  Pet food or groceries on the web—maybe this time?   On the other hand, we just invested in a great idea that goes back almost a century—more on that later.  But first, a couple of examples—from 1968 and 1994.

The Information Superhighway Started in…Orlando?

Sometimes a great idea fails to take hold the first time around because the technology doesn’t exist yet to enable it.  Today, we take for granted a virtually infinite selection of online entertainment, information, communications and ecommerce—but the original idea was revolutionary, and unachievable, less than twenty years ago.  I know, because I was there.

In 1994, Time Warner launched the Full Service Network (FSN), a groundbreaking interactive system that promised to give consumers an unlimited choice of home entertainment, information, communications and shopping, starting with 4,000 lucky households in Orlando, Florida.  Customers would cue movies and games on demand, shop the world’s stores from their armchairs, and order Domino’s pizza with a click of the remote control.

Despite the initial fanfare, the project failed due to its reliance on the television as the delivery device, massively expensive proprietary technology ($5,000 set-top boxes and a massive Silicon Graphics (SGI) server at every head-end), and the impossible challenge of inventing and integrating everything involved—interactive content, user interfaces, ordering and billing systems, and so on—from scratch.

In today’s world of Amazon, Netflix and Xbox, it’s funny to read quotes like this one from a 1994 New York Times article about the FSN (it’s worth reading the whole article to realize how far we’ve come):

There is great uncertainty about which services will be popular and whether they can be offered profitably, particularly exotic services, like interactive video shopping and grippingly realistic on-line video games.

It seems almost laughable now, but back then it didn’t seem so obvious.  As a VP at US WEST, one of the major players in Orlando and other projects, I ended up with responsibility for GOtv, an interactive TV guide to local movies and restaurants developed for this brave new world.  Orlando couch potatoes who actually wanted to go out would reserve a table or buy movie tickets in advance on GOtv – something like a cross between today’s Fandango and OpenTable.  At least that was the vision.  Like the FSN, GOtv was a great idea, but based on a flawed technology paradigm.

Of course, the idea of providing full interactivity and unlimited choice to consumers was a very sound one.  That same year, on the other side of the country, my now partner Marc Andreessen and SGI co-founder Jim Clark founded Netscape.  Within no time, Andreessen, Clark and a whole new generation of Internet entrepreneurs were leveraging affordable PCs, open standards and powerful network effects to create a massive ecosystem of providers and consumers and successfully deliver on the core idea of the Full Service Network.  (As for me, I redirected our visionary but flawed TV efforts to the emerging Internet, and soon entered the action directly by joining @Home Network, the pioneer of today’s high-speed Internet, in 1997.)

The Full Service Network: Great idea, wrong technology paradigm.

Loebel’s Blue Envelope

Unlike the Full Service Network, some ideas now succeeding phenomenally on the Internet also worked spectacularly the first time around, using the technology of the time.  In 1968, an entrepreneur named Terry Loebel invested $500 to mail a “cooperative envelope” with offers from 14 local businesses to 20,000 households in Clearwater, Florida (yes, Florida again!).  His sales staff made a name for themselves by zipping from merchant to merchant on roller skates.  The idea of mailing local offers to consumers proved highly compelling to small businesses.  Loebel’s startup, Valpak, grew explosively to become a marketing behemoth, ultimately mailing 20 billion offers a year in 500 million blue envelopes to 45 million households.

Forty years after Loebel’s first mailing, Chicago entrepreneurs Andrew Mason and Eric Lefkofsky founded Groupon, bringing the concept of targeted local offers to a global Internet market of 2 billion users.  Groupon uses email rather than snail mail, its salespeople connect with local businesses over the phone and Internet rather than on roller skates, and it’s pioneered breakthrough innovations like group buying and daily deals—connecting local businesses and consumers worldwide in ways Loebel could only have dreamed of.

Valpak:  Great idea, great first run, now back in a radically different incarnation for a spectacular second act.

Next up

That almost century-old idea I mentioned earlier?  That’s the topic of tomorrow’s post—and the inspiration for Andreessen Horowitz’s newest investment!