Jim Goetz and Jan Koum at Startup School SV 2014

Koum says moving from Ukraine to the United States at 16 made it hard to stay in touch with family and friends before internet messaging. At Yahoo, where he worked for nine years, he learned how to scale servers and recruited experienced engineers who later helped WhatsApp support a large user base with a small team.

Koum says he set out to build a product rather than plan to start a company. After buying an iPhone in 2009, he and Brian Acton first made a status indicator, then pivoted to messaging using contact and server code they had already built. Early revenue from the paid iPhone app helped them cover costs and choose investors carefully.

Koum says WhatsApp charged for its iOS app partly to slow growth while the team improved reliability, servers, and customer support. He describes avoiding press and meetings to protect time for product work. The team used phone numbers and address books instead of usernames, and supported platforms such as BlackBerry and Nokia early on.

Koum says a small server team, around seven or eight people, supported WhatsApp’s backend. He chose Erlang after failing to find a suitable open-source XMPP server, and says its message passing and clustering suited their needs. Goetz also recalls that at acquisition WhatsApp had about 20 to 25 engineers supporting roughly 400 million active users.

Koum says shared goals around connecting people and Facebook’s global outlook made the partnership a good fit. He says the discussions focused on continuing WhatsApp’s growth and mission, with him and Acton staying to lead the product. After the acquisition, the team remained focused on reliability, speed, backups, and phone or SIM changes.