3 awesome startups that failed
Share on Facebook
Share on Twitter
Share on Google+
Share on Reddit
Share on Email

Photo credit: Shutterstock, Gil C

Why being first often doesn’t work out

Tech in Asia

First-mover advantages are overrated. Look at Microsoft. Bill Gates was prescient, and they were onto the cloud, internet of things, and smartwatches since the early-2000s. But Gates’ prophetic touch did not translate to groundbreaking success in those areas

Instead, Google dominates the cloud, the internet of things is still nascent, and Apple is the frontrunner in the wearables race. In the fast-moving tech world, today’s champions quickly become tomorrow’s has-beens. Being the first mover is not the goal. Rather, startups should aim to be the last mover.

Here are three startups that were the kings of the hill at one point, but quickly lost out in the global dominance stakes.

Mt. Gox

At one point, Tokyo-based Mt. Gox was truly on top of the Bitcoin world. The online exchange, which converted fiat currency into Bitcoin, had 80 percent market share in 2013 and traded as much as US $25 million a day. That was before Bitcoin fever reached its highest point.

Photo credit: Shutterstock 360b

Photo credit: Shutterstock 360b

Why it failed: gross incompetence

In February 2014, hackers stole US $460 million from its coffers. Overnight, Mt. Gox crumbled. Yet the destruction was years in the making, according to this penetrating Wired story.

The company was managed poorly. The head of Mt. Gox, Mark Karpeles, was an engineer who was uncomfortable being a CEO. He was incompetent and apathetic. The hackers took advantage of a flaw in Bitcoin called “transaction malleability”. But Mt. Gox could’ve easily patched it up with its own software if it cared to. It didn’t.

Mt. Gox spent money on flat-screen TVs, $400 lunches, and a US$1 million cafe. It didn’t use any version-control system for its codebase, nor did it introduce a test environment until it was too late. As a result, a flawed code could introduce a vulnerability to the system, and the team wouldn’t realize it. Mt. Gox even flouted US regulations, resulting in federal agents seizing US$5 million from the company.

The result was disaster, not only for the company, but the Bitcoin world. Bitcoin prices fell as Mt. Gox’s troubles made the news. It reinforced the perception that the industry is run by shady characters who couldn’t be trusted. On the bright side, it signaled the end of a phase for Bitcoin, from a get-rich scheme into an era that will hopefully be marked by true innovation.

Friendster

Ah, good ol’ Friendster. Remember when leaving a testimonial on “the wall” was a thing?

At its height, the site had 115 million registered users. It got a decent acquisition offer from Google for US$30 million. It had a nice slate of investors, including venture capital firm KPCB and LinkedIn co-founder Reid Hoffman.

Why it failed: technical problems

Friendster founder Jonathan Abrams knew exactly where he wanted to take the site. In an interview with Mashable, he knew that Facebook could threaten Friendster’s dominance. He planned to build a news feed, a social graph, and an apps ecosystem on top of the social network.

He wanted to attract users in the colleges to stymie Facebook momentum among young adults.

But technological challenges slowed him down. Mashable wrote that “investors weren’t focused on patching up the service properly, let alone responding to the threats posed by upstarts like Facebook.” It became so bad that users had trouble logging into the website for two years, and load times were terribly slow. They failed to build their infrastructure to match the traffic. This allowed competitors Myspace and Facebook to catch up.

Facebook, on the other hand, abided by the adage “move fast and break things.” It’s an approach tailored to its gameplan: social networks succeed due to network effects, which means the more users you acquire in the fastest time, the better your chances of becoming dominant. A solid infrastructure allows Facebook to develop features and acquire users instead of worrying about server load, and adding more zeroes to user count is what matters most.

Friendster, of course, is still around today. But it’s unrecognizable from its former self. The site was acquired by Malaysian online payments company MOL for a rumored US$100 million in 2009. It then morphed into just another gaming portal on the internet.

Eachnet

Each-who? Don’t blame yourself if you don’t recognize the name. It was the leading online auction site in China with 90 percent market share, and got acquired by eBay for a cool US $150 million in 2003. That set the stage for a titanic struggle between the US giant and Alibaba.

Why it failed: failure to localize

Photo credit: company website

Photo credit: company website

Eachnet was a moderate success story, especially for its original founders. According to the new book Alibaba’s World, it already had a dominant market share in China’s online auction market by the time eBay first invested in the company in 2002.

Alibaba meanwhile was succeeding in the B2B space. It then worked on its consumer marketplace Taobao in secret to stop eBay’s expansion in China. When the time came to unwrap Taobao, the company launched a media blitz against eBay Eachnet.

In a war of words, it focused on a key differentiation: Taobao was entirely free to use, while eBay charged sellers a commission and product listing fees. Alibaba founder Jack Ma made a bet that eBay wouldn’t match its move due to shareholder pressure back home. It’s not just a marketing stunt: the Chinese don’t trust new online services easily, and they’d rather try something first before paying for it. Taobao was focused on giving them a taste.

Taobao also differed from eBay in not using the auction model. While American consumers had tons of collectibles lying around the house ready to be sold online, the Chinese didn’t – after decades of poverty and upheaval, people were only just becoming consumers and aspiring to move up to the middle class. Because of that, Taobao focused on offering a huge variety of goods to buyers.

Another major difference stemmed from Taobao’s superior understanding of its local market. In China, people are accustomed to building relationships before conducting business. eBay was opposed to that because allowing consumers to talk to one another negated their business model of collecting commissions. Because Taobao was free, it encouraged communications by building a live chat feature in the site for that purpose.

Taobao’s cultural knowledge impacted the way it did payments. eBay relied on PayPal, which allowed direct transactions between buyer and seller. This was untenable for Chinese consumers, who don’t trust each other. Meanwhile, Alibaba’s payment system Alipay operated on an escrow model, withholding payment to the seller until the buyer has verified the product has arrived and is in good condition.

eBay Eachnet also made a few strategic mistakes that sealed its fate. It froze local website development in favor of a global site, believing that consumer tastes in China would follow those from developed countries. It got rid of animations and cute icons, which the Chinese actually preferred over a minimalist design. It prioritized international users and forced the Chinese to register new accounts, which meant they lost income.

Finally, by hosting the site outside of China, it became excruciatingly slow for Chinese users. As in the case with Friendster, a slow site leads to a lot of frustration for users, causing them to jump into the embrace of a nimble competitor.

“eBay may be a shark in the ocean, but I am a crocodile in the Yangtze River,” Alibaba founder Jack Ma famously told Forbes. “If we fight in the ocean, we lose – but if we fight in the river, we win.”

This article originally appeared in Tech in Asia

Share on:Share
Share on Facebook
Share on Twitter
Share on Google+
Share on Reddit
Share on Email
Terence Lee

About Terence Lee


Based in Singapore, Terence writes about technology trend and startups in Asia. Passionate about harnessing tech for storytelling, he has picked up web development, sniffing his way around Ruby on Rails, Javascript, and D3.js. His articles have been published on Venturebeat, Yahoo!, SGE, Straits Times, and Today. He digs movies, computer games, and food porn.

More Goodies From Industry


Socially-focused startups tackle rural Mexico’s energy problems

Endless Lima traffic spawns innovative startups

New concept: Booking meeting rooms at the heart of Tel Aviv by the hour