Meta Sells Attention

Thesis

Meta sells attention. Facebook, Instagram, and the rest of the Family of Apps are inventory. Advertisers buy the inventory. Reality Labs is a long, expensive bet that the next inventory will be a headset. Platform policy is how the inventory stays sellable.

Context

I have an Instagram account and a WhatsApp install because other people already do. I do not publish mystic-bytes into a Meta feed as a growth strategy. The writing lives on a site I host. I use the apps the way a remote person in Auckland uses them in 2026: messages, a few accounts I actually wanted, a lot of ranking I did not ask for. I have never worked at Meta. The useful documents are the 10-K, the Community Standards, and the public wreckage of the old Platform API.

Calling these products “communities” is a category error. Communities do not run a ranking function whose job is dwell time. Advertisers do not buy communities. They buy impressions next to whatever the ranker decided you would not close.

Mechanism

The feed is a ranking problem with an ads invoice. Meta’s annual report still describes a Family of Apps funded overwhelmingly by advertising.¹ That is the business. The mechanism is a model that predicts what you will linger on, then inserts a paid slot. Reels, Stories, the old News Feed: different chrome, same job. Engagement is not a vibe. It is the objective function. When people say the apps “got worse,” they often mean the ranker got better at its actual assignment. I have closed Instagram after twenty minutes and been unable to name a single post I meant to open. That is not a personal failing. That is a successful session from the seller’s side.

WhatsApp is the graph that does not look like a feed. Messaging is utility. Utility is how you stay installed when the feed feels extractive. Meta bought WhatsApp in 2014 and spent years arguing it was a separate culture. Business accounts, status, channels, and the slow arrival of ads around the edges are how a messenger becomes inventory without admitting it in the first release notes. I keep WhatsApp because people in other timezones keep WhatsApp. That is a lock-in that does not need a like button.

Reality Labs is future inventory, billed as research. The segment loses money in public filings, year after year, at a scale that would kill a company that was only a headset shop. Meta can afford it because ads print the cash. Quest and the Horizon language are a bet that the next attention surface will be owned the way the phone feed is owned. I have not bought a Quest. I can still read a loss-making segment as strategy: if the phone ranker saturates, buy the face.

Platform policy is the control surface. Community Standards, ad policies, and API lockdown after Cambridge Analytica are not a civics class. They are how a company that sells attention survives regulators, brand safety teams, and its own history of letting third parties vacuum the social graph.² When Instagram throttled the old API, developers called it a betrayal. It was a portfolio decision: the graph is too valuable to rent cheaply. I do not need a conspiracy to explain a closed API. I need an incentive. Attention you cannot meter is attention you cannot sell.

You are not the customer unless you bought an ad. Creator funds and “professional dashboards” complicate the sentence without reversing it. The advertiser is the customer. The user is the inventory with a login. I try to remember that when an app asks for more notifications. Notifications are a collection mechanism.

Tradeoffs

Reach vs being the product. If you need to find a person who only lives on Instagram, you will open Instagram. That is rational. It is also how the ranker gets another session. Prefer a link to your own site. Use the app as a pointer, not as a CMS.

Messenger utility vs family gravity. WhatsApp is the hardest to leave because it is the address book. Leaving Facebook is easy if you never posted. Leaving the messenger is a group project. Budget the group project, or admit you are staying for other people’s defaults.

When the apps are the right tool. Event coordination, a family chat, a business that already bought the ads: use them as tools. Turn off the feed if you can. I treat Meta like a casino I enter for a specific table. I do not owe the floor a wander.

Close

Attention is the SKU. Ads are the payment. Policy is the fence. Reality Labs is the sequel if the phone saturates. I will keep the messenger I cannot replace and keep mystic-bytes off the ranker. Call the company by what it sells. It sells you, in slices, to people with a media budget.

— JV · Dark Heart Labs.

References

Meta Platforms, Inc., Form 10-K. Family of Apps advertising versus Reality Labs segment results. The invoice and the long bet, in the form the SEC requires. ↩
Meta, Community Standards and Business / ads policies; Facebook Platform API restrictions after 2018. Policy as inventory control, not as a town square charter. ↩