The Network That Lived on Desktops

What happened

The company was Facebook. In 2012 it reversed its approach to mobile almost completely. It rebuilt its iPhone app natively, releasing the new version in August, and followed with Android. Mark Zuckerberg, its chief executive, said at the TechCrunch Disrupt conference that September that betting too much on HTML5 rather than native apps had been one of the company’s biggest strategic mistakes, and that it had cost about two years. He described Facebook as now a mobile company, and said that mobile users were more likely to use it daily than desktop users.

The organisation changed with the product. Facebook moved from functional groups to product groups, each responsible for its product on every device, and dissolved the separate mobile team into them; a product director described the aim as making every engineer a mobile engineer.

The ads moved into the feed. Promoted posts inside the mobile news feed, already running before the IPO, were reported to draw about twice the clicks of desktop ads and nearly three times the price, and were grossing about $500,000 a day by July 2012. In October Facebook opened mobile app install ads, which let developers pay to promote their apps in the feed, to all developers.

The IPO itself went badly. Shares priced at $38 on 18 May 2012 fell to about $17.55 by early September, as investors worried about mobile. Then the numbers turned. Mobile accounted for about 23 percent of Facebook’s advertising revenue in the fourth quarter of 2012, and about 53 percent a year later, when mobile ad revenue passed $1.25 billion in a single quarter.

The alternative of owning a layer of the phone was tried too, and failed. In April 2013 Facebook launched Home, software that replaced an Android phone’s home screen, along with a phone made by HTC. Reviews were poor, sales of the phone were reported to be very small, and the effort was quietly wound down. By 2025 Meta, as the company was renamed, had revenue of $201 billion, almost all of it from advertising, and 3.58 billion people a day using at least one of its apps, nearly all of them on phones.

Judging the decision apart from the outcome

The process was strong in the way it confronted evidence. The company’s own data showed users moving to phones, its own prospectus named the risk, and its own engineers knew the apps were poor. Rather than defending the earlier web-technology decision, which would have been the path of sunk costs and pride, the leadership reversed it publicly within months. It also changed the organisation, not just the product, recognising that a mobile team inside a desktop company would always be second, a lesson familiar from the innovator’s dilemma.

The decision was made under the pressure of a public listing, which made it harder. A company focused on the quarterly revenue that analysts watched might have protected the desktop column ads; Facebook accepted a period in which its share price halved. The willingness to bear that cost depended partly on the founder’s voting control, which shielded the decision from investors’ short-term reactions. That shield was a condition of the decision’s quality, and it is not available to most companies.

The decision also had costs that the revenue figures do not show. Putting ads inside a personal feed on a device carried everywhere, and building products for daily use on that device, intensified the competition for attention that later drew criticism from users, researchers and regulators. Those consequences were partly foreseeable in 2012, and they belong in any full judgement, though they do not change the commercial verdict.

What a good process would have weighed

The web-first path had genuine merit, independence from platform owners and a single codebase, but the evidence from users said it was not working, and the company could not afford to wait for the technology to improve. The own-platform path addressed a real strategic weakness, dependence on the operating-system companies, and Facebook’s later failure with Home suggests users did not want a social network to become their phone. Protecting the desktop was the comfortable choice, and the history of platform shifts suggested it was the most dangerous. The chosen path accepted short-term pain, a public admission of error and a more intrusive ad format in exchange for following users where they had already gone.

Sources and limitations

The numbers on users and the risk to revenue come from Facebook’s own prospectus. Zuckerberg’s statements are from a public interview after the decisions were made, and describe his own judgement of the earlier strategy. The figures on the early performance of mobile feed ads come from a third-party agency study and press reports and were not disclosed by Facebook. The sales of the HTC phone were reported by a single unnamed source. The point about the founder’s control of votes is the author’s interpretation of why the decision could withstand the fall in the share price, as is the judgement of the decision’s wider costs.