He Built a Hundred-Person Company Chronicling the Boom. Now He Chronicles Startups' Decline | What Is Silicon Valley Anxious About ❺
A data company that grew up the old-fashioned way — and a close observer of how American tech is changing.
A data company that grew up the old-fashioned way — and a close observer of how American tech is changing.
From: Anand Sanwal Date: February 24, 2016, 18:24:51 Subject: F’ed Company
Body: Those of you old enough to remember the 1999-2001 dot-com bubble might recall an obscure website called f’ed company, which chronicled the trivia of the bubble years. …… In the unicorn era, we launched the Downround Tracker, following the companies that failed to keep the myth going……
Love you. Sent lovingly from the bathroom, Anand
CB Insights email, February 24, 2016
In 2001, after the dot-com bubble burst, 56% of companies didn't survive a year. So someone built a website called Fucked Company to track the startups about to go under.
Fifteen years later, a crop of richly valued "unicorns" can't go public, their valuations are being cut, and some are being bought on the cheap... So Anand Sanwal built a similar data service to track valuations:
Not long ago, Sanwal sat down for an interview with QDaily.
For this Indian-American entrepreneur — a man whose mind wanders, who likes to crack jokes, but who always brings the talk back to the heart of the question — tracking failing companies is only a side project.
Sanwal is the founder of the data company CB Insights. Started in New York six years ago, it now has 97 employees.
Like most data companies, CB Insights makes its money selling expensive industry reports — $40,000 apiece and up — and even more expensive consulting.
But what made it famous is the free data newsletter Sanwal writes every day. Leave your email address on the CB Insights site and six days a week you'll get a message with a subject line like these:
- "Peter Thiel Loves Drinking Young People's Blood": a startup trying to ease Parkinson's symptoms by transfusing patients with young people's blood, which Peter Thiel was said to be considering investing in;
- "So-Called PaaS Is Really Pizza as a Service": mocking what PaaS actually stands for, "Platform as a Service";
- "Fat-Skinny Startups": poking fun at the Lean Startup idea Silicon Valley loves to champion;
- "You're Fired!": meant to mock one company's habit of firing people, but the subject line frightened plenty of office workers, who thought it came from their boss.
Every email carries jokes like these, along with straight-faced charts reading the tech trends. And instead of closing the way English letters usually do — "Thanks," "Best regards" — Sanwal signs off "I love you."
Honestly, this kind of clickbait isn't far from the headlines on WeChat official accounts.
The difference is that Sanwal's emails rest on a deep pile of data. Just about every publicly disclosed startup investment seems to turn up here.
Over the past two years, CB Insights has become the investment data firm American media quote most often.
John Markoff, the New York Times's most senior technology reporter, wrote recently in a story about the AI boom:
According to the market research firm CB Insights, funding for artificial-intelligence startups quadrupled over the four years from 2011 to 2015, reaching $681 million. The firm also expects AI startups to raise $1.2 billion this year, up 76% from last year.
During the wave of unicorn valuation cuts that began in 2015, CB Insights was busy supplying data too. Its charts and figures have become the authority people reach for whenever the startup market comes up.
But when Sanwal set out eight years ago, he never imagined the company would end up like this.
Sanwal's business started with an "internet thinking" idea that flopped
"At first we were called Chubby Brain," he told QDaily.
He also told us the story behind the name CB Insights. This was 2008:
"When we showed the product to people at investment banks, they said, 'We love your data product, but there's no way we're buying a service called Chubby Brain.' So, later, we changed it to a more grown-up name: CB Insights."
Sanwal laughed a little at himself: the "more grown-up" name was just the initials of those two words, plus "Insights" — a word big companies love.
Chubby Brain came from a domain name Sanwal had registered.
Back when the internet was still a novelty, Sanwal, fresh out of Wharton, made pocket money buying and selling domains. He bought more than 200 of them.
In 2001, after the bubble, he joined American Express to work on investments and acquisitions.
By 2008, when he decided to leave and start something of his own, Sanwal had made vice president at the high-end credit card company. His job was to invest $50 million of the company's money — to find the startups that might threaten American Express's future business, and to invest in or even buy them early.
The idea for the company also came from a problem he hit while investing. Sanwal says he disliked the data products on the market: expensive and useless.
"My boss asked me to find some solid startups in the payments industry. My first thought was that this should be easy — the kind of easy where you click the mouse a few times and there it is."
So he searched the databases his company paid for — Dow Jones, Thomson, S&P — and found that each one gave different numbers. Then he tried Google and couldn't find much either.
"The company pays all that money for these databases and the products still don't work. I didn't want to sit in a meeting looking stupid and get blamed for 'missing this company.' I'd already been thinking about doing something on my own, and that's when I thought: why not build a better solution myself."
2008 was a lively year for startups. The opening of the mobile internet war showed people how much was possible:
Apple's iPhone turned one and its App Store went live; countless startups began their businesses there.
Google, three-plus years after its IPO, had put investors' doubts to rest, and its share price had climbed back to the highs of a year or two earlier. The search giant everyone once suspected of being overvalued was now the stock everyone chased. Meanwhile its base in the mobile market, Android, went on sale with HTC's G1.
Many of the startups we know today began in 2008. Uber's two founders, unable to get a cab in Paris, were angry enough to start a company. The home-rental service Airbnb, not long out of the startup incubator Y Combinator, was looking for its first round.
Social networks were in their best years too: Mark Zuckerberg's founding story had just been made into the film The Social Network. Twitter caught fire with the 2008 presidential election.
Sanwal and CB Insights's other founder, Jonathan Sherry, wanted to make something with the internet as well. So they built a site that was half Wikipedia, half review platform, devoted to reviewing companies that hadn't yet gone public or been acquired. Anyone could submit an account of what was really going on inside a company.
Next to today's CB Insights, that first idea was more tech, more startup, more "internet thinking."
Little more than a year later, Sanwal wanted to shut the product down. Nobody used it. "People had no reason to review these companies, and I'm not even sure whether they just weren't interested."
Still, the wiki database gave them some hope of a pivot. Sanwal noticed that the infographics they published on funding and acquisitions drew crawlers from other sites, and that investors were writing in to ask for the raw Excel files behind the data and how they'd managed to present it that way.
"So we just put up a paywall. That started in 2010, at $2,400 a year."
What actually took off was doing it by hand
"A lot of what we do is pretty manual. The experience may not apply to other companies, but for us this approach worked," Sanwal said.
Since he wasn't a programmer, Sanwal started out searching the web by hand and typing what he found into spreadsheets.
"The first 50,000 data points really were entered by hand, by me and Jonathan. We'd go to news sites, find an item like 'So-and-so just got $30 million from Sequoia,' write it into the spreadsheet, and then repeat that 49,999 times," Sanwal said, laughing.
Once they had programmers, CB Insights began using search-engine-style crawlers to pull information off websites — investment firms' own pages, company press releases, news sites, tech blogs...
Early on, to draw paying customers, Sanwal wrote the occasional free data newsletter, linking to the company's recent reports.
Well, at the start those emails looked much like plenty of other failed newsletters:
A year of them brought CB Insights just over 1,400 subscribers — many WeChat official accounts add more followers than that in a day or two.
"We used to think an email for people in the industry had to be dull to count..." Sanwal recalled.
Sanwal's first data report covered venture capital in 2010. He knew other data and consulting firms would publish something similar, so his team finished a week early, added their own reading of the numbers, and sent it to reporters. ReadWriteWeb decided to use the data from that CB Insights report.
"The media were our first customers, more or less," Sanwal said — and the team contacted every one of them by hand.
"Say we want to do a report on artificial intelligence now. We go online, see which reporters have written about it, and find their public email addresses," Sanwal said. At the beginning they knew no one in the press, but every time they planned a report, they asked reporters in advance whether they were interested; if so, they sent it early, or even put together a custom data report for that outlet.
"Custom data reports are the most important work we do!" Several years on, CB Insights has more than 500 journalists it contacts regularly — and Sanwal says this "find the reporters by hand" method continues.
In the two hottest years for startups and investing, Sanwal's "WeChat official account"-style emails took off
The real change began in 2013. Sanwal was following the data blog of the American dating site OKCupid and the data section of the political news site FiveThirtyEight. He found these data pieces, with a point of view behind them, a lot of fun, and decided to push his own newsletter that way.
"Once we changed the style, our subscriptions jumped."
CB Insights also caught a good moment. In 2013 there was another important shift in tech — Facebook changed how the capital markets saw startups.
Facebook had long drawn doubts: a $100 billion valuation and not much profit. Before its 2012 IPO, Facebook's valuation was cut by a third at one point; the day after it listed, the stock fell below the offer price, and the deal was called the worst American IPO in a decade.
A year later, with ad revenue up sharply and more than a billion active users, Facebook's market value was twice its pre-IPO valuation.
And this was exactly the business CB Insights wanted to be in.
To attract more potential paying customers, the free data newsletter became one of the company's most important jobs.
"On average I spend half an hour a day getting the email written." He had finished that day's just before our interview.
"Sometimes it's hard, because on top of the data you have to come up with a good joke. But on the whole, readers' feedback has been good." As of now CB Insights's email has more than 180,000 subscribers, and keeps adding a thousand-plus a week.
By the figures Sanwal gave earlier, the emails have a 25-30% open rate, and 3.5% of readers write back to ask what the paid service costs.
In the rush, Sanwal has missed the "I love you" sign-off a few times, and then got teasing replies: "So you don't love us anymore?"
As more people signed up for the free email, the price of CB Insights's paid data rose with it. Today it costs at least $40,000 a year to buy the basic CB Insights data service.
"What we do is use data to predict — to find the next market, the field where the next acquisition or investment should happen, what the next move by a company in the same field might be, all in a very data-driven way."
"Some people pay us anywhere from $100,000 to $200,000, and we give them the financials of different startups, scores for how active they are on social media, and there are more data- and algorithm-based products in development." Sanwal and his team don't plan to raise money; they want to earn it from paying users. In 2015, CB Insights took in more than $1 million from them.
When startups hit real trouble, old-line companies become the buyers of last resort
By 2016, though, the weather had changed again.
Startups are also finding it hard to cash out by going public. By Fortune magazine's count, American tech companies that listed in 2015 earned, on average, negative $9 million. And the "unicorn" tech companies that listed before them have fallen further still:
We also see more and more companies with high valuations that keep losing money and don't go public. Worse, some of these unicorns have built the best product in their industry and still can't show results to match the valuation.
Since the end of 2015, at least half of the 30 most highly valued startups in Silicon Valley have gone through a valuation cut, an exodus of employees, or something worse.
That is why Sanwal built the Downround Tracker this year. Data, he believes, can do more than tell people what is going well; it can also give a clear-eyed view of the market's rise and fall.
The newest trend is that startups unable to cash out through an IPO would now rather sell the company. And compared with the big tech companies, big traditional brands are more inclined to buy and invest in technology startups to stay competitive. In the past month alone there were three large acquisitions:
- August 16: Ford bought the Israeli machine-vision startup SAIPS, preparing to build self-driving cars;
- August 9: Walmart bought the e-commerce company Jet.com for $3 billion, hoping to take on Amazon;
- July 20: Unilever bought the men's razor subscription service Dollar Shave Club for $1 billion;
A little earlier, Nike, Adidas and ASICS each bought a startup to build fitness apps.
To say nothing of the traditional carmakers buying up startups in bulk to get into ride-hailing and self-driving:
What they are doing is what Sanwal did at American Express.
"Big companies face risk — they worry a small company will take their business away. So the fund we ran then mostly invested in companies that might become the next American Express, and sometimes in our own projects."
"The biggest change in these years is that every industry has started to worry about technology." As conditions shifted, startups with nowhere else to go have gradually sold themselves to traditional companies.
After eight years following the startup boom, Sanwal's customers have turned back into traditional companies like American Express — just as they were when he first started investing, right after the dot-com bubble burst.
Editor's note: PC, internet, smartphone — the spread of every foundational technology has come with years of fast growth. For the tech industry, that growth has lasted forty years.
Now it may pause for a while: growth from smartphones has slowed, and after ride-hailing and food delivery we haven't seen another breakthrough on the order of social networking or e-commerce, the kind that shifts a hundred billion in wealth.
What new technology will pave the way for the industry's next stretch of growth? And will it arrive in time?
These questions will affect everyone, from giants like Apple and Google, to venture capital firms and startups, to every user of every tech product — basically everyone.
Over the past forty years, most of the technologies that laid the foundation for the tech industry's long boom came out of Silicon Valley. We will report a series from Silicon Valley, looking at this change from different angles:
In Silicon Valley, we saw what really happened to the "WeChat official accounts" of four years ago
Translated from the original QDaily article. Original publication date retained.