The transformation from Facebook to Meta is not merely a name change — it carries a significant message about the structural shift of the business, restoring investor confidence, and betting its future on artificial intelligence (AI). Meta now clearly reveals itself as a personal-centric platform and AI company that has moved beyond being a "social network." Through recent earnings, internal statements, and the evolution of its advertising business, we trace the context of this transformation step by step.

1. The "Meta" Story Begins Again, and an Earnings Reversal

The author has often linked the "here we go again" meme with Meta (formerly Facebook), but this time, he can no longer use it. That's how completely Meta has entered a new phase.

In the recently announced Q2 2025 earnings, Meta recorded an impressive 22% revenue growth, and its bold investment in AI was also received positively by the market, alongside the solidity of its advertising business.

"Meta Platforms reported 22% revenue growth in Q2, demonstrating that its core advertising business remains strong even as it invests billions in AI."

Net income also exceeded market expectations at $18.3 billion, and with up to 24% revenue growth expected in coming quarters, investors showed a more positive than anxious attitude toward Meta's "AI investments."

2. The 2021-2022 "Metaverse Shock" and Its Lessons

In 2021-2022, Meta's major news revolved around the company name change, massive metaverse investment, and concerns about growth stagnation. Meta's stock price plummeted to below $100, and the author himself covered the collapse of investor confidence in a piece called "Meta Myths" at the time.

Behind this rebound lies not just a simple reversal but important lessons and strategic changes. An anecdote from CFO Susan Li about an investor meeting in October 2022, shared on a podcast, is particularly striking.

"I get that you're building the future of computing, the next mobile platform. But why should I own your stock right now? Can't I just invest after the results come in a few years from now?"

Susan Li says this question left a deep impression. Through this, Meta's leadership came to painfully realize that you can't earn investor trust by emphasizing only "the future" — you must also consistently demonstrate "current performance."

3. The Investor Trust Model: Amazon vs. Old Meta

In the past, the archetypal company that attracted investment based on a distant future vision was Amazon. Even when it seemed far from profitability, trust in its infrastructure and platform investments persisted.

"Amazon's shareholders are satisfied with being part of a powerful consumer benefit engine rather than expecting fat profits."

Over time, Amazon reaped high returns from AWS and other ventures, and investors could clearly see the connection. But Meta never received similar trust and always faced skeptical eyes.

"Myspace still exists, and it was the instinct of investors to anxiously wait for the emergence of a new social network that would replace Facebook."

But now, with a strategy that demonstrates both "investment for the future" and "current performance," investor sentiment has completely shifted.

4. Changes in the Advertising Business and AI's Real Impact

Looking inside Meta's advertising business this quarter, AI is at its core.

Unlike previous quarters, the increase in ad impressions was notably larger, and the decline in unit prices was modest, resulting in excellent overall ad performance.

"The strong ad performance this quarter is largely thanks to AI. AI-powered recommendation systems boosted ad efficiency, with approximately 5% more ad conversions on Instagram and 3% more on Facebook."

An important point here:

• Meta's AI achievements primarily come from recommendation systems (Andromeda, GEM, Lattice, etc.), representing a substantive contribution distinct from simple LLM (large language model) investment.
• The author also points out that investors have become too generous with the word "AI," failing to distinguish between existing product improvements and next-generation AI (superintelligence) research.

"The results we're seeing now are thanks to 'improvements to already-existing core products,' not the effects of next-generation superintelligence investment."

5. Reels, Ad Inventory, and the Company's Growth Dial

The growth of ad inventory from new products like Reels has been a major contributor to Meta's advertising growth.

• 2017: Instagram feed ad saturation → Stories ad launch triggers impression surge
• 2020: Stories saturation → Impressions rise again with the full-scale launch of Reels
• 2024-2025: Reels also saturating, new growth drivers needed

The author notes that recently, there appeared to be no clear new inventory creation factor, but in reality, "increased video consumption time" and "ad load optimization" (i.e., showing more ads) were the secrets behind the impression surge.

"The increase in impressions was primarily driven by growing video consumption on Facebook and Instagram, and additionally by ad load optimization."

In other words, as users repeatedly watch more short videos, ad exposure opportunities naturally increased. Meta can now adjust its "growth dial" as needed.

6. Platform Identity and the Shift to "Personalization"

At the root of all these changes is the shift in Meta's identity from a "social network" to a "personalized content platform."

As recently as 2016, Zuckerberg championed "connecting people" as the company's core mission.

"Over the next 10 years, we'll continue to invest in platforms and technologies that connect more people and places. The three pillars of our innovative long-term roadmap are connectivity, artificial intelligence, and virtual and augmented reality."

But in recent earnings announcements, the word "connection" barely appears once, replaced instead by a focus on "individual empowerment."

"Meta's vision is to provide everyone with a personal superintelligence. Through this, each person can leverage AI for whatever they find valuable."

Even in legal defense arguments, Meta mentions that it is no longer a friend-based social network.

"Today, the social element across all Meta services — the time spent consuming content from friends — is just 7% on Instagram and 17% on Facebook. Most of it is watching short videos recommended by AI, which is the result of competition with TikTok and YouTube."

It is now clear that "personalization" has been re-established as Meta's core value, replacing "groups, communities, and connections."

In Closing

Meta (formerly Facebook) has evolved beyond its identity as a social network company into an AI-based personalization platform and a new business engine. At the center of this are the immediate results of its advertising business, practical AI applications, and the restoration of trust with investors. Meta's growth now unfolds on a new trajectory of "optimizing individual experiences" and "maximizing ad inventory," rather than expanding connections. 🚀