To Sell or Not to Sell

Meta’s ambitious AI data center expansion is prompting a critical question for CEO Mark Zuckerberg: should the social media giant monetize its burgeoning compute capacity by entering the cloud infrastructure market, or prioritize retaining it for its own aggressive artificial intelligence development? While Meta’s capital expenditures in AI infrastructure rival those of established hyperscalers, it remains the sole major player without a dedicated cloud services business. However, Zuckerberg has recently begun to explore the possibility of launching such a venture, a move that could leverage its substantial computing power in a market currently facing significant resource constraints.

Industry buzz suggests this strategic pivot is already gaining traction, with reports indicating that AI firm Anthropic is in preliminary discussions with Meta to lease computing power. Zuckerberg himself commented on the lucrative demand during Meta’s recent second-quarter earnings call. “We’re getting a lot of offers for compute at a significant premium over what we paid for it,” he stated, also hinting at ongoing development of enhanced coding and productivity tools.

These comments surfaced concurrently with Meta issuing a less-than-robust revenue forecast for the third quarter and reporting a stark 90% year-over-year decline in free cash flow, primarily attributed to soaring capital expenditures. The market reacted with concern, pushing Meta’s stock down over 7% in after-hours trading, adding to an 11% decline year-to-date.

Adding to the expenditure narrative, Meta has revised its 2026 capital expenditure guidance upward, raising the lower end to between $130 billion and $145 billion. This aggressive investment strategy mirrors that of its hyperscaler peers. Just last week, Alphabet increased its guidance to $205 billion and notably experienced negative cash flow for the first time. Microsoft also projected approximately $175 billion in capital expenditures for the year. Amazon is expected to report its results shortly.

Investors have been keenly awaiting clarity on Zuckerberg’s AI strategy, which has been perceived by some as somewhat unfocused, potentially leaving Meta trailing behind competitors like OpenAI, Anthropic, and Google in the race for cutting-edge AI models and services. Analysts have voiced the need for a clearer vision. “I think everyone wants clarity into what he wants to do in the compute business,” noted Brent Thill, an analyst at Jefferies.

### Navigating the Compute Conundrum

While Zuckerberg provided limited specifics regarding his cloud ambitions, he elaborated on the strategic considerations involved. “In terms of running the business, obviously, a common trade-off that we need to make is around how much do you monetize something today versus develop future assets,” Zuckerberg explained. “I think that it’s always a portfolio.” He further elaborated that a potential enterprise business wouldn’t solely focus on selling raw compute capacity, but could also encompass offering API access, productivity services, and Meta’s internally developed AI agents. “And I think that there’s just a very, very large opportunity there,” he added.

However, Meta’s own burgeoning AI ambitions necessitate substantial compute resources, particularly as the company, under the leadership of AI chief Alexandr Wang, aggressively rolls out new models. This month, Meta unveiled its Muse Spark 1.1 model, which Wang described as the “strongest model for agentic and coding work yet” and at a competitive price point compared to offerings from OpenAI and Anthropic. Zuckerberg stressed the importance of retaining sufficient internal capacity, stating, “It would be foolish to basically just sell all of the compute and take a short-term profit.”

Acknowledging the challenges of entering the enterprise sector, where Meta has historically faced hurdles, Zuckerberg conceded that building a robust presence will require significant effort and the development of new skill sets. While not explicitly detailing hiring plans for a dedicated sales force, such a move would be critical for Meta to effectively penetrate the business market. “That’s going to be somewhat a new muscle that we build as a company,” Zuckerberg emphasized. “But I think it’s a very important one that we build.” Meta has recently secured a strategic hire in Dave Brown, a former senior executive at Amazon Web Services, signaling a serious intent to bolster its enterprise capabilities.

A significant hurdle for Zuckerberg in convincing Wall Street of his vision lies in Meta’s past strategic bets. The company’s extensive foray into the metaverse, initiated in 2021, continues to incur substantial quarterly losses. Meta’s Reality Labs division, dedicated to virtual reality devices and wearables, reported a deficit of $4.62 billion in the latest quarter, against a revenue of just $431 million.

Despite these challenges, Meta is driven by a compelling need to diversify its revenue streams beyond digital advertising, which currently accounts for a staggering 98% of its income, and to solidify its position as a key player in the AI landscape. Zuckerberg appears committed to this aggressive pursuit. “I get that this is sort of a big bet across the industry,” he concluded. “My personal bet is that the people who invest in this are going to be rewarded and feel very good over time.”