All posts in “Startups”

GitHub’s scandalized ex-CEO returns with Chatterbug

Translation earbuds might eliminate some utilitarian reasons to know a language, but if you want to understand jokes, read poetry, or fall in love in a foreign tongue, you’ll have to actually learn it. Unfortunately, products like Rosetta Stone leave people feeling burned after claiming the process should be easy while never helping you practice talking with a real native speaker. You know, the skill you actually want. Just memorizing vocabulary doesn’t make you fluent.

So after teaching millions of people to code better, a team of former GitHub co-founders and executives this week launched Chatterbug to combine the best of online and face-to-face foreign language learning. Starting with German, Chatterbug uses a homegrown video chat alternative to Skype that lets you simultaneously talk, type, read, and screenshare your way to becoming conversational.

But one of the co-founders’ past may cast a shadow over Chatterbug. Tom Preston-Werner resigned from his role as CEO and co-founder of GitHub following an investigation into allegations of harassment and intimidation of a female employee by he and his wife Theresa Preston-Werner.

GitHub employee Julie Horvath told TechCrunch that Theresa had bullied her about not writing negatively about the company, said she could read employees’ private chats and had spies at the startup, and verbally bullied her.

While an independent investigation claimed to have found no evidence of illegal behavior or gender-based harassment on Tom’s part, it did conclude that the former CEO showed “mistakes and errors of judgment” and “insensitivity to the impact of his spouse’s presence in the workplace and failure to enforce an agreement that his spouse should not work in the office.”

Ex-GitHub CEO and Chatterbug co-founder Tom Preston-Werner

We asked Tom how he’s building Chatterbug differently this time around. “With some hindsight, the organic management structures at GitHub were a double edged sword. It unleashed a lot of creativity, but was fragile in handling conflict” says Preston-Werner. “From the very beginning of Chatterbug I’ve had serious conversations with the other founders on how to use those experiences to create a more robust channel of communications.”

Former GitHub head of comms and Chatterbug co-founder Liz Clinkenbeard tells TechCrunch “In retrospect, I think one of the major challenges at GitHub back then was that the company’s fairly flat structure sometimes made it difficult to know who to talk to about problems, and how to resolve them before they escalated.” With Chatterbug, she says the team has “been very open and deliberate about wanting to foster a safe and supportive work environment.”

It’s possible that Tom’s inclusion on the team could make it tougher for Chatterbug to hire talent, especially women. Though at least it seems the company is taking office demeanor and harassment issues seriously as it grows.

“I’ve always tried my best to empower my teammates and create a work environment that every employee will love. I haven’t been perfect at that endeavor in the past” admits Preston-Werner. “But I’ve learned much from those experiences and intend to use that knowledge to ensure that Chatterbug is a safe, welcoming, and productive place to work for women and other folks traditionally underrepresented in the tech industry.”

Cutting Skype Out Of Language Learning

Scott Chacon discovered what was broken about the current crop of language learning tools when he tried to pick up French via Duolingo and Japanese through Skype chats before spending time in the two countries. “I realized there was a gap between the digital apps that are super flexible but aren’t very effective at teaching conversation with real people, and the tutoring systems or in-person schools that were inflexible and super difficult to do” Chacon tells TechCrunch.

So he started building his own tools that would blossom into Chatterbug. The former GitHub co-founder and CIO recruited GitHub’s Clinkenbeard, director of engineering Russell Belfer, and Preston-Werner over late 2015 and early 2016. They raised a $1.8 million seed round from SV Angel and Berlin’s Fly Ventures to have early-stage allies on both sides of the pond.

Setting goals in Chatterbug

Now after some private trials starting in March, Chatterbug just launched the public beta of its German learning program, with Spanish and French coming next. And right out of the gate, it’s trying to set reasonable expectations for how fast people can pick up a new tongue. “The most difficult part of being in the business is that Rosetta Stone and other companies try to sell the idea that language learning can be easy” Chacon says. “Learning a language is not easy. It’s like a marathon.”

That’s why one of the first things you do in Chatterbug is adjust a slider for when you want to be fluent by, and it tells you how frequently you’ll have to study and be tutored. The app then gives you a foundation of vocabulary using “spaced repetition”, a study method employed by medical students where questions you get wrong get shown more often while you’re displayed fewer questions like those you got right.

Chatterbug understands when you almost get an answer right

Then Chatterbug schedules you for one-on-one tutoring over its video chat system designed specifically for language learning. Rather than having to commit to a weekly session time, only learn when your particular tutor is available, or fall behind if you miss a group class, you just punch in when you want to practice. Chatterbug pairs you with whatever appropriate tutor is available, gets them up to speed on your progress, and provides a personalized curriculum of exercises to do together based on what you’ve been screwing up.

The heavy engineering background of the Chatterbug team allowed it to create a WebRTC-based video chat that lets you view files together with your tutor and see each other’s cursors as well as talk and type. That’s a huge improvement over trying to pass PDFs back and forth or figure out what exercise the teacher is discussing.

Chatterbug’s video chat lets you talk, type, view files, and see each other’s cursors

The pricing model flexes to accommodate your pace. You can get all the self-study features plus one live lesson a month for €15 or eight for €80 with extra sessions costing €12 each if you want to take a vacation next year. Or for €195 you get unlimited sessions and can learn a language in just a few months. Chatterbug is also going B2B, appealing to businesses trying to educate employees by offering discounts and easy expensing.

Turning Anyone Into A Teacher

Chatterbug co-founder Liz Clinkenbeard

The startup’s data-driven approach could make it quick to expand to more languages and identify what’s toughest to learn. Chatterbug gives you the option to have it store recordings of your video sessions, and even give it permission to use them for research. Clinkenbeard studied linguistics at Harvard, and is using her expertise to help the company determine what are the most common vocab and grammar mistakes to help you avoid them.

Long-term, turning native speakers into tutors could offer new employment options to those lacking other quantifiable skills. “After leaving GitHub, I wanted my next project to be something that would positively impact a lot of people. As a filter, I’d ask myself ‘could this idea lead to the creation of a million jobs?’” says Preston-Werner.

Chatterbug faces a wide range of competitors like Rosetta Stone, Duolingo, Busoo, Babbel, and HelloTalk — some with deep pockets and a penchant for downplaying the difficulty of reaching fluency. Being real with people doesn’t always make for great marketing, and people who failed with other products exhibit a “healthy amount of skepticism” says Clinkenbeard. Then there’s the looming threat of advancing translation technology, like the new Google auto-translating Pixel Buds headphones.

Still, “I don’t think it will destroy the need for language learning” says Chacon. “At some point, in-person translators will be obsolete. Not sure if that’s in 5 years or 45 years.” But even if we solve information translation, culture translation will still be in demand. “You don’t want to wear an ear bud while you’re getting married” he laughs. At a time when the world is increasingly polarized and xenophobic, understanding your fellow humans without a technological intermediary could generate some much-needed empathy.

Sneaker and streetwear reseller Stadium Goods just launched their first app


Stadium Goods, the online (and brick and mortar) marketplace for highly sought after sneakers and streetwear, is launching its first app.

Live today to coincide with the startup’s two year anniversary, the first iteration of the app is basically just a mobile marketplace.

But Stadium Goods plans to eventually build out this functionality and take advantage of location services and push notifications to get creative with its marketing strategy, the startup explained to TechCrunch. They also plan on giving app customers early access to specific product launches, which is a tactic that’s proven successful with other streetwear retailers.

While the sneaker resale market is crowded with big venture-backed startups like StockX and Goat, Stadium Goods is the only one wasn’t launched around a mobile app. It’s also the only one with a brick and mortar strategy. The startup has two locations in New York, which account for about 10% of the company’s total sales, with the other 90% happening online.

And Stadium Goods explained that out of this 90% of transactions occurring online, the “majority of them” are already happening via mobile. This should translate into a strong adaptation rate as customer migrate from shopping on the mobile web to their native app.

Stadium Goods has raised about $5.6M since launching two years ago. To coincide with their app launch they’re also running a 20% off sale on all purchases today.

You can check out their app on iOS here and on Android here.

The next tech bubble will last weeks, not years, and it will be gloriously insane

I’ve been there for the early ICO (initial coin offering) days for cryptocurrencies. I’ve seen companies raise hundreds of millions of dollars within hours, and I’ve seen valuations for startups that don’t even have a working product climb to billions. 

And now that these clay giants have begun to crumble, I look back at the beginnings of the ICO craze, and ponder when did it all go wrong. 

Well, it couldn’t have been more than a few months ago, because this entire bubble has been going on for about a year. 

Perhaps this is the real gift the blockchain technology, which is the basis for most of these companies: Speed. The ability to raise money fast, to build a product quickly, to burn brightly and, in some cases, die fast. 

Don’t believe me? Ethereum, the platform on top of which most of these ICOs are based, was launched July 2015. It took two years from that to Vegas clubs adopting the technology for cheaper drinks and stripper tips

I’m not saying the ICO bubble has burst. The initial coin offerings — digital token-based fundraisers for (mostly) blockchain-based startups — are still going strong, with dozens of new ones lined up every month. More than a billiion dollars have been raised so far, but that’s a paltry sum compared to the trillions raised (and lost) in the dot-com craze in the early aughts. The craze will likely go on for a while, but the early signs of decline are there: ICOs for companies that don’t do anything; companies that raised hundreds of millions squabbling over what to do; authorities clamping down on scams. 

I’ve been there every step of the way, writing about this trend and even investing in some ICOs, and the thing that impressed me the most is the insane pace at which everything is happening. 

ICO funding in Q2 2017.

ICO funding in Q2 2017.

Image: Coindexk

During the dot-com days, in the nineties, the money had flown into crappy companies at an unprecedented rate. Think of a web-based project, build a .com website, raise money, poof: You’re rich. 

But there were still some regulatory hurdles to overcome. You had to talk to lawyers, banks, VC funds. Yeah, the pace was crazy — building an online store was way faster than building a brick and mortar one. But it still took roughly seven years for the bubble to peak and burst. 

Seven years is forever in ICO land. The thing that was hot yesterday is dead today and forgotten tomorrow. 

Just look at Kik’s Kin, one of the largest, most high-profile ICOs. A month ago, the company raised nearly $100 million in its token sale. But right around that time the tide turned and most investors decided that any ICO (unless it has vast, game-changing potential, which Kin doesn’t) that tries to raise over $50 or so million is too greedy and is not worth investing in. Right now, Kik tokens are trading at less than half of their ICO price, which was unimaginable just two months ago, when nearly every high-profile new token skyrocketed in value immediately after it hit the market. (Disclosure: I participated in this ICO and I own Kin tokens). 

And does anyone still remember Hubii, the token that Floyd ‘Crypto’ Mayweather shilled in August?  Despite a $2.7 million market cap, the token is being traded in mere thousands of dollars in daily volume, and from what I can see on social channels, near-zero buzz. 

And the next big thing? It literally changes from hour to hour. A few days ago, when the price of Bitcoin surged due to the promise of free money after the two upcoming forks, people sold ICO tokens en masse, with many proclaiming the golden era of ICOs dead on Slack channels and social media. Now that Bitcoin’s price is retreating, ICOs are hot again, but only those with a small market cap, a solid team, a sound whitepaper, at least two advisors, and a big amount of a largely intangible asset called hype. Don’t have the hype? You’re as good as dead. 

I’m not sure what lies ahead for ICOs; no one is. There will probably be more regulation, and there will surely be less space for crappy startups to raise millions, both of which are good things. While it does indeed look like the golden age is behind us, perhaps there is more growth ahead — sound growth, with better foundations. 

But following this world has taught me that when the next bubble comes along, it will blow up as fast as the hype around the new episode of Stranger Things. There will be no time to read books on investing or learn candlestick charting techniques. Perhaps even Slack, Telegram, and Reddit, currently the main hype-producing social channels, will be obsolete. If you want in, you won’t be able to afford to stick with what you know; you’ll have to jump head first and learn as you go along. 

I’m not sure whether the next bubble will have anything to do with blockchain, but this technology definitely sped up the pace at which tech revolutions can happen. Just like the dot-com boom in the nineties, when so much business moved from offline to online, the blockchain (and, largely, Ethereum) have enabled startups to arise quickly, get funded quickly, and solve important problems in the way we do business. Make no mistake: Next time around, things will happen even faster. 

It’s both fascinating and scary, and I can’t wait for it to happen. 

Disclosure: The author of this text owns, or has recently owned, a number of cryptocurrencies, including BTC and ETH. 

Studio makes running more exciting with coaching, music and competition


Jason Baptiste wants you to run.

He said he’s seen the personal benefits of running since 2009, when he became unhappy with his weight and committed to run a “daily 5k,” something he’s held to ever since.

“Not only has running helped me be healthier, it’s helped my mind and spirit become stronger,” Baptiste wrote. “Running has become an outlet for me to be a better person.”

Despite his personal connection, Baptiste might not seem like the most obvious person to launch a new fitness startup — his last company, Onswipe, was a mobile publishing startup acquired by Beanstock Media in 2014.

But he told me, “Group fitness classes are media businesses” — they’re all about bringing an audience together to watch a central performance (with, okay, a lot of audience participation).

Similarly, Peloton has found success with live streamed spinning classes. In fact, you can think of Baptiste’s new startup Studio as an attempt to offer a Peloton-style service for running, without selling you the actual exercise equipment.

Studio has created iPhone and Apple Watch apps with a variety of running classes, combining coaching and music. Baptiste said this is designed less for experienced runners and more for newbies who want the health benefits but maybe see the running as painful and boring (which, to be fair, it totally can be). So Studio is all about “turning fitness into entertainment,” while also making sure you get a good workout.

[embedded content]

Studio-Teaser from Studio on Vimeo.

There’s a competitive aspect too, as Studio awards you “Fitcoin” for the time and distance of your runs. (You’ll need the Apple Watch app to track your runs and earn Fitcoin.) The virtual currency gets you a ranking on the Studio leaderboard and can also be used to earn prizes.

The app is designed specifically for treadmill runners, with treadmill-specific instructions on how to adjust your speed. Baptiste said he’s interested in supporting “running beyond the treadmill” in the future, but he also thinks he’s addressing a huge group already, as treadmills are the biggest seller among exercise equipment.

Studio is available for download now. There’s a subscription price of $15 a month or $99 a year.

What washing dishes, driving a truck and working in a cemetery taught me about the power of ‘blue-collar’ software


I work as a venture capitalist in a glass-walled office and wear dress shirts to work. But it wasn’t so long ago that I wore an orange neon vest instead. That was when I stocked shelves at a big-box, home-improvement chain. I also drove a truck — where I was once nearly arrested for incorrectly filling out my vehicular paperwork on the job — washed dishes, waited tables, was a telemarketer and even hauled dirt at a cemetery.

I don’t think this is the usual route to a job in tech and venture capital. But my former work life has opened my eyes to a corner of the technology world I think has the chance to create the next Salesforce, Oracle or LinkedIn: software targeted at workers, often blue-collar, who do their jobs outside corporate offices.

The forgotten workers

For those of us sitting at our desks and working behind laptops on programs like Microsoft Office, it can be easy to overlook the large, sometimes forgotten, workforce out there in construction, manufacturing, transportation, hospitality, retail and many other multi-billion-dollar industries. Indeed, more than 60 percent of U.S. workers — and even more globally — fall into these “blue-collar” industries.

By and large, these workers have not benefited much from recent technology improvements available to office-based workers — think new email and workplace-collaboration technologies, or advanced sales and HR systems. Never mind the long-term opportunities from technologies like artificial intelligence, drones and virtual or augmented reality for companies in these sectors; hourly and field workers are dealing with much more basic on-the-job challenges, like finding work, getting their jobs done on time and getting paid.

These more basic needs can be solved with seemingly simple technologies — software for billing, scheduling, navigation and many other business workflows. These kinds of technologies, unlike AI, don’t automate away workers. Instead, they empower them to be more efficient and productive.

Today, these technologies for hourly and blue-collar workers are finally proliferating and coming into their own. This is mainly because of two broad technological shifts: the rise of the smartphone and the advent of cheaper and easy-to-deploy, customizable “cloud” software.

These new technology products can, for example, help plumbers view new jobs, navigate to their next appointment and bill customers on the fly. My firm recently backed a company based in Southern California that does just that. Another one of our recent investments, also in this area, lets maintenance and facility workers snap a photo of a broken piece of equipment on a manufacturing line and create and route a work order — all in seconds on their phone. In the past, most of this was done on paper, or via very clunky, desktop software.

The market for this type of technology is potentially vast. Given that today’s publicly traded cloud-software companies have created roughly $200 billion in market value by targeting the 40 percent or so of the population that work in offices, I believe that the opportunity to build software for blue-collar workers is just as large.

In part, I believe in these markets because I’ve been doing “diligence” on them for nearly 15 years.

In venture capital, being too early is the same as being wrong

After holding my various odd jobs in cemeteries, big-box stores and more, I began my professional career in a part of the tech industry that wasn’t nearly as glamorous as going to work for Goldman Sachs, McKinsey or Microsoft. Building on my earlier experience, I started work at a company building a software business targeting plumbers and heating and air conditioning and electrical contractors.

The idea was to provide customers with a mobile app that would enable them to do things like receive work orders and manage their day-to-day operations and invoices in the field — all from their mobile phones and without having to return to the office. The only challenge was that the basic technology infrastructure to make this type of solution viable just wasn’t available yet. These were pre-iPhone/Android days. My business was stuck trying to sell expensive, ruggedized hardware that cost between $2,000 and $3,000 per bulky phone before we could sell any software. Coupled with data-connectivity issues in remote areas, getting the business to take off was an uphill battle.

Today, thanks to better mobile technology and the cloud, the landscape is completely different. Several companies are capitalizing on these trends and providing specialized software to workers globally. Procore, for example, serves the construction general-contractor market, providing software for tablets that GCs can carry around construction sites to view blueprints and plan tasks. And a number of other companies — among them Australia’s Deputy, Sweden’s Quinyx and When I Work, based in Minnesota — make scheduling software to replace manual time sheets for hourly workers in industries like restaurants, healthcare and many more.

Times have clearly changed. In today’s smartphone-centric world, blue-collar software is now, in my opinion, a good value and potentially transformative for many companies. I may have hung up my orange neon vest, but I’m a believer in this technology.

Featured Image: Martin Barraud/Getty Images