Why eBay Isn’t Big in Japan: Learning from a Billion-Dollar Mistake

In part two of a three-part series on software scalability, Mike Fisher shares a three-step process aimed at helping tech companies prepare for and properly manage the challenges of rapid expansion and growth.

In the summer of 1999, online auction site eBay was preparing to scale its platform internationally, setting its sights specifically on Japan, a market that was virtually begging for its own Japanese-language auction site.
Accomplishing that goal should have been relatively easy, too. After all, eBay was the dominant auction site in the United States, so translating its technology and scaling it for the Japanese market should have been fairly straightforward.
That’s not exactly how it worked out, however.
By 2001, Businessweek Magazine reported that eBay owned just 3 percent of a market in which $1.6 billion in goods were traded. Yahoo! Japan, on the other hand, owned 95 percent. What caused eBay to lose so much ground to a competitor that was better known domestically for being a search engine and content hub?
“The bottom line is that eBay hadn’t fully prepared for the myriad of challenges that scale can present, and that caused some issues,” says Mike Fisher, veteran technology executive and former PayPal VP of Engineering and Architecture. “Ultimately, eBay delayed its deployment in Japan by six months because of those issues, which gave Yahoo! the opening it needed.”
Unfortunately, that unpreparedness cost eBay several billion dollars in market cap, and the company has yet to displace Yahoo! in Japan.

Managing Growth and Scalability the Right Way: 3 Steps

So, with the benefit of hindsight, what might eBay have done differently?
According to Fisher, who co-founded AKF Partners in 2008 to help companies better manage scalability, the most important lesson that growing technology businesses can take away from eBay’s misstep is that scalability is a perpetual, progressive challenge — not one that can simply be addressed in times of crisis.
“You can’t just wait for something to go wrong to start addressing scalability,” Fisher explains. “That ad-hoc approach can lead to missed opportunities, because if something goes wrong and you aren’t prepared for it, that opens the door to competing solutions or products.”
To avoid that fate, Fisher says companies need to take three key steps over the course of two years in order to successfully manage the ups and downs of software scalability:

1. Design: Pick a value of scale (i.e. a specific revenue or growth target, new market, new product release, etc.) that your company expects to reach within two years. What systems architecture issues will you face, and which parts of your system will be vulnerable as it scales? The goal here is to identify potential roadblocks to scale and design around them about 24 months before they are actually needed.
2. Implement: After you know how you need to design to scale, Fisher says the next step is to actually implement the processes that will get you to where you want to go. This could include writing code, ensuring that existing code can be read from one database and write to another, or splitting customers across multiple databases. The idea is to implement the changes long before you actually plan to need the scale (ideally, at least 12 months out). 
3. Deploy: With about three months to go before needing to scale, Fisher says companies should deploy the code to specific hardware so that they’re fully prepared to go live. In cloud environments like AWS or Rackspace, this deployment can take place much closer to the actual user demand. The goal is to not spend money on hardware and server space way before you actually need it, but instead ramp up in a way that ensures you aren’t running behind schedule.

“Inevitably, things may change over the course of two years and the business goals you were designing for initially might not be relevant when you’re ready to deploy,” Fisher points out. “But the general idea here is to think very strategically about why and how you’re going to scale, and what issues might crop up as you execute that process.”

Scalability is a Marathon, Not a Sprint

Ultimately, Fisher says it’s most important for companies to remember that scaling is a journey, not a spur of the moment initiative.
“As long as your company is growing, you need to be working on scalability projects,” Fisher explains. “The scope of those projects will vary depending on your development stage, of course, but it’s something that you need to be constantly aware of.”

What are the biggest challenges your business has encountered along the path to scale? How are you successfully navigating around them and managing growth?