What the business is
Udacity sells online 'nanodegrees' in technical subjects such as AI, self-driving cars and digital marketing to consumers and enterprises.
The bet
Cutting complexity, slashing content costs and giving every student a paid mentor can lift graduation rates and make Udacity profitable.
How it started
Udacity's 2017 revenue grew 100% on programs like its self-driving car and deep learning nanodegrees; 2018 then fizzled. CEO Vishal Makhijani left in October and co-founder Sebastian Thrun stepped in as interim CEO and head of content.
What happened
About four weeks before the article, Thrun sent an all-hands email saying the trajectory was not long-term tenable and that 'we must act, and act now'. About 130 people were laid off, the content team shrank from about 140 to 40, and a former HP and GE executive, Lalit Singh, joined as interim COO. A mentor program started, matching every new student with a mentor.
What has to be true
The 2015 unicorn round of $105M set growth expectations the 2018 business could not meet
Thrun judged the company had grown too quickly and become burdened with red tape
High-polish video production created a bottleneck in how many courses could be released
Personal mentors were the lever he believed could raise graduation rates and unit economics
What can be applied
A founder returning to a company that grew too fast can restructure quickly by cutting production cost and shifting spend to student outcomes, but the results are untested at the time.
Aftermath
As of 22 February 2019 the mentor program and new leadership were weeks old; Thrun himself said it remained to be seen whether the strategy would work, and unnamed sources inside the company voiced doubts.
FOLLOW THE EVIDENCE