Platforms that pick your clothes, find the right people
By Dennis Clemente
Meetup showcases platforms that shops for you, picks right people for the job
NEW YORK—The AXA Equitable Center makes for a grand entrance. Thus said Matt Turck, long-time host of Data Driven, as he welcomed the crowd to its majestic auditorium, complete with velvet curtains and flattering spotlight. One of the most attended meetups in the city, Data Driven is holding its meetups at AXA for a few months until the Bloomberg auditorium finishes its renovation.
At the meetup last March 16, Data Driven divided the talks based on its format. Eric Colson, chief algorithms officer of Stitch Fix; and Kieran Snyder, founder & CEO at Textio presented their companies while Peter Fenton, general partner at Benchmark and Eliot Horowitz, co-founder & CTO at MongoDB sat with Turck to discuss their companies and their industry in general.
Colson opened the night’s data talk with StitchFix. “There is no shopping in our site, because people hate shopping.” That got people’s attention. What StitchFix does is create your style profile and give you five hand-picked items. You keep what you like and send the rest back.
Recommendations have worked for several companies. For Amazon’s sales, 35 percent; Linkedin’s connection’s, 50 percent; Netflix’s watched movies, 75 percent and StitchFix, 100 percent of its sold merchandise.
Colson said they combine both data and human insights to make StitchFix work. There’s no denying the importance of human insights because of their wealth of experience, according to him. “(But) they can’t be doing the same things as (its data/algorithms),” he said.
Next presenter Snyder said Textio mines data from recruiters and hiring mangers to find patterns that work, showing how it works to help companies hire better. It was as simple as copy pasting a job posting from a site to a Textio blank field.
Using statistics and machine learning, it analyzes job text and outcomes data using listings from a set of companies. It makes use of patterns that it finds to predict the performance of job postings and help you fix it before you ever publish it, with analytics and feedback right as you’re typing. It makes use of color to highlight words (green for phrases that work) and red (for least successful ones) that should help its clients get the talents they need.
It offers real-time feedback as well as sharing and collaboration on job listings with colleagues. On average, Snyder claims that people who use Textio see a 24-percent lift in qualified applicants, a 17-percent drop in time to hire; and a 12-percent increase in underrepresented applicants. “We found words like synergy don’t work among underrepresented applicants,” she said.
Snyder said the best feedback loop comes from its customers, as she also observed how job listings can amplify a company’s voice, throwing wrong assumptions about the lack of creativity of job listings. Expedia is one of its customers.
With MongoDB, Horowitz asked the audience, how many are frustrated with their databases? When MongoDB came into the picture in 2007, it was tackling what is seemingly a persistent problem with databases. In 2009, Salesforce ported their database to MongoDB.
“Developers say (MongoDB) is a pretty good experience,” says Horowitz, adding it looks forward to making users more productive by offering more ways for developers to keep using it.
Addressing a monetization question, Horowitz offered consulting and support, its BI tools and cloud services.
Started in 2007, MongoDB takes pride in having 85 percent of work done in New York. In 2015, the company released its 3.2 version that helps address a pernickety issue these days—encryption. It also started a BI connector with Tableau and Compass.
What does it take to be an entrepreneur? Peter Fenton, who invested in Twitter when it was only 25 people, echoed the sentiments of Paul Graham of Y Combinator: “Is the entrepreneur deeply authentic?” He also points out how feeling uneasy can actually work for you, if he means being grounded enough to think of the realities of the startup business.
“Take two those variables and layer around that,” he said.
As for figuring out which is promotional and authentic among the current crop of startups, he describes the tech startup world based on how whales breach and then submerge again. “We’re moving (in a) cycle, but we’re making the ecosystem healthier.”
Fenton pointed how institutional money may have given tech startups longer capital runway and burn rate, but valuations do go down and money may not be as easy to get.
For radical growth, Fenton thinks ubiquity is crucial. However, he points out how some technology has a gestation period (before they hit critical mass).
