How freewheeling Twitter became a money-spinning juggernaut

Updated 02 October 2013
0

How freewheeling Twitter became a money-spinning juggernaut

AROUND MIDNIGHT on Christmas Eve of 2009, a handful of employees at Twitter received an unconventional holiday greeting from Dick Costolo, then the chief operating officer.
“It was an e-mail that said, ‘We have to move really, really fast. There’s no time to rest because we have a massive opportunity in front of us,” recalled Anamitra Banerji, who headed the team that built Twitter’s first advertising product. “It was kind of crazy because we were all on break, but that attitude was exactly what we needed at Twitter.”
The company is now on the verge of fulfilling the opportunity Costolo foresaw as it prepares for the most highly anticipated initial public offering since Facebook’s debut last May. The offering is expected to value Twitter at up to $15 billion and make its early investors, including Costolo, very wealthy indeed.
Yet Twitter’s quick transformation from an undisciplined, money-losing start-up into a digital media powerhouse took every bit of whip-cracking that Costolo could muster, along with a rapid series of product and personnel decisions that proved effective even as they disappointed some of the service’s early enthusiasts.
Costolo was a comparative late-comer at Twitter, joining the company three years after it’s 2006 launch, but the company increasingly bears his imprint as it hurtles toward the IPO: Deliberate in decision-making but aggressive in execution, savvy in its public relations and yet laser-focused on financial results.
Costolo has not flinched in pruning and reshaping his management team, while Twitter, the company, has been ruthless in cutting off the smaller companies that were once a part of its orbit. A one-time comic actor who cut his teeth in business at Andersen Consulting before starting several companies, Costolo may never be as closely associated with Twitter as Mark Zuckerberg is with Facebook, yet he is arguably just as important.
“The founders consider Dick a co-founder, that’s how deep the connection is,” said Bijan Sabet, an investor at Spark Capital and a Twitter board member from 2008 to 2011. “He’s not this hired gun to run the company. He understands building out the business but also the product, strategy, vision.”
Twitter declined to make Costolo available for comment, citing the pre-IPO quiet period.
When Twitter’s then-CEO Evan Williams brought on Costolo, an old friend and colleague from Google Inc, as COO in September of 2009, the three-year old company was already under pressure.
The microblogging service was gaining hip, young users at an unprecedented pace, and its trio of co-founders — Williams, Biz Stone and Jack Dorsey — had been splashed across magazine covers as the embodiment of San Francisco cool. Yet the whispers in Silicon Valley were growing louder: Twitter didn’t have the technical chops to make the service reliable at huge scale, and it didn’t have any way to make money.
“Having been on the core original team of engineers, we didn’t have the skills among us to build a world class service,” said Alex Payne, an early Twitter engineer, noting that many of the team members came from smaller start-ups and non-profit organizations rather than established Web giants like Google.
Williams viewed fixing the site’s notorious technical problems as the top priority but was ambivalent about the business strategy. For months, people familiar with the situation say, Williams weighed options ranging from display advertising to licensing Twitter’s data to becoming an e-commerce hub to offering paid “commercial” accounts to businesses.
Costolo — who had sold Feedburner, an advertising-based blog publishing service he founded, to Google for $100 million — had no such doubts. By his second month on the job, he had helped persuade Williams to green-light engineering positions to build Twitter’s first ad unit, which would become the “promoted tweet” — the cornerstone of Twitter’s business today.
“Dick’s conversations with Ev were key,” said Banerji, now an investor at Foundation Capital. “He had a fundamental belief that this was the future of Twitter monetization and said, ‘You have to do it.’“
Over four months in early 2010, Costolo, working closely with Banerji and Ashish Goel, a Stanford engineering professor who specialized in the science of auction algorithms, to refine the promoted tweet. It resembled an ordinary Twitter message in every way, except that advertisers could pay for it to appear at the top of users’ Tweet streams and search results.
Costolo threw his heft within the company behind the advertising strategy. In early 2010, as the ads team drew up a related product called “promoted trends,” Costolo privately told them to make sure he was in the room when they pitched the product to Williams, so it would get pushed through.

A central mechanism governing the promoted tweet was “resonance,” a concept coined by Goel. Because Twitter users can re-circulate or reply to tweets, including paid advertisements, the company had the real-time ability to gauge, which ads were most popular, and those ads could then be made more prominent. And because the ads appeared in the same format as other tweets, they were perfectly suited to mobile devices, which could not handily display traditional banner ads.
When Costolo unveiled the promoted tweet in April 2010, Twitter announced it as a trial for only five brands, including Starbucks Corp. and Virgin America, and users almost never saw the ads.
But by the summer of 2010, Costolo felt confident enough in his concept that he began seeking a deputy to ramp up the company’s sales effort. For months, he courted Adam Bain, a rising star at News Corp, and at the same time began assiduously courting marketers, from corner suites on Madison Avenue to industry conferences on the French Riviera.
“Hashtags,” which help people find the conversations they’re looking for on Twitter, soon grew ubiquitous on TV, appearing in Super Bowl commercials, at Nascar races and on the Oscars red carpet.
“It wasn’t easy for Twitter to explain to people why they should buy content on Twitter until they sold it as a companion to TV,” Ian Schafer, the chief executive of Deep Focus, a digital advertising agency. “Now you’re even seeing the networks selling Twitter’s inventory for them. That’s magic.”


Tesla in Autopilot sped up before Utah crash

Updated 25 May 2018
0

Tesla in Autopilot sped up before Utah crash

  • Heather Lommatzsch, the driver of the vehicle, told police she thought the vehicle’s automatic emergency braking system would detect traffic and stop before the car hit another vehicle.
  • Police say car data show Lommatzsch did not touch the steering wheel for 80 seconds before the crash. She told police she was looking at her phone at the time and comparing different routes to her des
SALT LAKE CITY, US: A Tesla that crashed while in Autopilot mode in Utah this month accelerated in the seconds before it smashed into a stopped firetruck, according to a police report obtained by The Associated Press Thursday. Two people were injured.
Data from the Model S electric vehicle show it picked up speed for 3.5 seconds shortly before crashing into a stopped firetruck in suburban Salt Lake City, the report said. The driver manually hit the brakes a fraction of a second before impact.
Police suggested that the car was following another vehicle and dropped its speed to 55 mph to match the leading vehicle. They say the leading vehicle then likely changed lanes and the Tesla automatically sped up to its preset of 60 mph (97 kph) without noticing the stopped cars ahead of it.
The police report, which was obtained through an open records request, provides detail about the vehicle’s actions immediately before the May 11 crash and the driver’s familiarity with its system.
The driver of the vehicle, Heather Lommatzsch, 29, told police she thought the vehicle’s automatic emergency braking system would detect traffic and stop before the car hit another vehicle.
She said she had owned the car for two years and used the semi-autonomous Autopilot feature on all sorts of roadways, including on the Utah highway where she crashed, according to the report.
Lommatzsch said the car did not provide any audio or visual warnings before the crash. A witness told police she did not see signs the car illuminate its brake lights or swerve to avoid the truck ahead of it.
Lommatzsch did not return a voicemail Thursday. A Tesla spokesperson did not respond to a request for comment.
The car company has said it repeatedly warns drivers to stay alert, keep their hands on the wheel and maintain control of their vehicle at all times while using the Autopilot system.
Police say car data show Lommatzsch did not touch the steering wheel for 80 seconds before the crash. She told police she was looking at her phone at the time and comparing different routes to her destination.
She broke her foot in the crash and this week was charged with a misdemeanor traffic citation. Online court records do not show an attorney listed for her.
The driver of the firetruck told police he had injuries consistent with whiplash but did not go to a hospital.
Tesla’s Autopilot system uses cameras, ultrasonic sensors and radar to sense the vehicle’s surrounding environment and perform basic functions automatically.
Among those functions is automatic emergency braking, which the company says on its website is designed “to detect objects that the car may impact and applies the brakes accordingly.” Tesla says the system is not designed to avoid a collision and warns drivers not to rely on it entirely.
The National Highway Traffic Safety Administration has said it is investigating the May 11 crash.
Tesla’s Autopilot has been the subject of previous scrutiny following other crashes involving the vehicles.
In March, a driver was killed when a Model X with Autopilot engaged hit a barrier while traveling at “freeway speed” in California. NHTSA and the National Transportation Safety Board are investigating that case.
This week, Tesla said Autopilot was not engaged when a Model S veered off a road and plunged into a pond outside San Francisco, killing the driver.
Earlier in May, the NTSB opened a probe into an accident in which a Model S caught fire after crashing into a wall at a high speed in Florida. Two 18-year-olds were trapped and died in the blaze. The agency has said it does not expect Autopilot to be a focus in that investigation.