Pepper Salé: Lessons from the bitter Aldebaran / SoftBank project



This guest post is authored by Mark Bivens. Mark is a Silicon Valley native and former entrepreneur, having started three companies before “turning to the dark side of VC.” He is a venture capitalist that travels between Paris and Tokyo (aka the RudeVC). You can read more on his blog at or follow him @markbivens. The Japanese translation of this article is available here.


Bloomberg published an article last week on the failed Pepper program, the emotionally intelligent robot originally developed by French company Aldebaran Robotics which was largely acquired in 2012 by Japan’s SoftBank for approximately $100 million.

The demise of Pepper has long been rumored (in fact The Rude Baguette originally broke the story nearly two years ago), though only recently is it becoming acknowledged more openly in Japan. On a recent visit to a business meeting in Tokyo, I was warmly welcomed at the interphone (despite my poorly accented Japanese), directed into a elevator, escorted to a conference room and offered coffee, all by a charmingly hospitable Pepper receptionist.

Although the Rude Baguette post‘s comments go deeper in speculation and entertaining mud-slinging, one part of the Bloomberg piece in particular resonated with me:
Aldebaran and SoftBank’s cultures didn’t mesh well. Engineers in Japan fumed when their French counterparts disappeared for weeks on vacation.

Aldebaran employees, accustomed to a flat structure, suddenly found many of their decisions second-guessed by an army of managers in Tokyo.

The Japanese parent created SoftBank Robotics Corp. to oversee the business and sell Pepper. It named Fumihide Tomizawa, a business manager who doesn’t speak English or French, to oversee development. Son put Takashi Tsutsui, a close ally and a veteran network engineer, in charge of technology.

So here are a few lessons from the unfortunate and probably avoidable demise of Pepper.

Acquisitions don’t magically solve organizational dysfunction. I understand from some of the firm’s early investors that Aldebaran faced internal lack of cohesion in the period prior to the SoftBank’s rescue, with allegedly some investors reaching the limits of their patience with Aldebaran management. Cleaning up personnel challenges in French companies is already difficult enough for shareholders locally, let alone a foreign acquirer.

Cross-cultural training is critical in transcontinental acquisitions like this one. The company cultures of SoftBank and Aldebaran fell in stark contrast: a hierarchical organization vs. a flat one; structured vs. chaotic; a work environment where office face time is valued vs. one in which the minimum 5 weeks vacation plus RTT days are prized.

Employees with an affinity to bridge the cultural divide — be it thanks to their languages skills, international experience, etc. — should be empowered to play a prominent role in such partnerships. Even among those who are not viewed as sufficiently senior in the hierarchy or as sufficiently expert in the domain, any individuals who can help facilitate the interpersonal connections are valuable during the initial critical phase of a merger.

Finally, perhaps the Aldebaran / SoftBank experiment shines light on a need that efforts like #LaFrenchTech might address: with its access to a pool of resources who are versed in political relations and diplomacy, a government initiative like #LaFrenchTech could provide guidance to startups in finding their way in unfamiliar territory.