Part 3 of After the Feed, a field guide to how social media gets rebuilt. See all parts →
Part 2 ended on a claim: the opening for a new social platform is authenticity, and the overlooked capability is the phone’s ability to prove something is real. That could stay a nice theory. Except it isn’t only a theory — pieces of it are already shipping, making money, and keeping people around. They just don’t look like “the next social network.” They look like small, odd, single-purpose apps.
Put four of them side by side and they stop looking random.
Locket: the constraint is the product
Locket does one thing: it puts a photo from a close friend directly on your home screen, as a widget, no app to open. And it enforces a hard cap — you can only share with a small handful of people.
That cap is the whole idea. There’s no follower count, no public feed, no algorithm deciding who sees what, no audience to perform for. By removing the things every other social app maximizes, Locket became genuinely intimate — and it worked commercially, racking up on the order of 80 million downloads and steady subscription revenue (Sensor Tower pegged a recent month around $300k), after raising $12.5M back in 2022.
The proof point: stripping the audience out of “social” isn’t a compromise. For a lot of people it’s the feature they were missing.
Retro: people will pay not to be hooked
Retro, built by former Instagram engineers at Lone Palm Labs, is almost aggressively calm. It’s a chronological photo journal you share only with friends, with weekly recaps instead of an endless feed. No ads. And it deliberately reintroduces friction that the industry spent fifteen years removing — you can mail an actual physical postcard of a photo from inside the app, for a couple of dollars.
Here’s the part that should get any builder’s attention: people pay for this. Retro runs on subscriptions — around $36 a year — and in 2026 it raised a $21 million Series A (led by Field Ventures and Thrive Capital) at a valuation north of $100 million, with roughly 1.7 million monthly users and in-app spending up more than 460% over half a year.
The proof point: a non-addictive, ad-free, deliberately slow product isn’t charity. It’s a fundable business, precisely because it refuses the feed model.
Halide: the anti-AI camera
Your phone doesn’t really take photos anymore; it computes them. Every shot is fused from multiple frames, denoised, sharpened, and smoothed by AI before you ever see it. Halide’s Process Zero exists to turn all of that off — a single exposure, straight from the sensor, no computational photography, no neural retouching, delivered as a plain HEIF and a raw DNG. It even runs on non-Pro iPhones, and it’s faster than Apple’s own raw pipeline.
People buy a camera app whose entire selling point is doing less to the image. The photo looks like light hitting a sensor, not like a model’s idea of a nicer version of your face.
The proof point: there’s real, paying demand for optical honesty — images you can trust because the machine kept its hands off them. That’s the authenticity wedge from Part 2, in shipping form.
Instax: the object that can’t be faked
Zoom out from apps entirely. Fujifilm’s Instax instant cameras — physical film, a print you hold — quietly became one of the company’s crown jewels, with the business hitting record sales around ¥150 billion (over a billion dollars), led largely by Gen Z and millennial buyers. In an era of infinite, editable, AI-generated images, a chemical print that developed in front of you is the one photo nobody can quietly alter later.
The proof point: the appetite for tangible, un-fakeable artifacts is large, growing, and generational — not a nostalgia blip.
What the four have in common
None of these is “the” replacement for the feed. Individually they’re a widget, a journal, a camera mode, a toy. But read together, they’re a map of the demand:
- Constraint beats infinity (Locket’s cap, Retro’s slowness).
- Authenticity beats polish (Halide’s raw honesty).
- Tangibility beats ephemerality (Retro’s postcards, Instax’s prints).
- Subscriptions and objects beat advertising as a way to get paid.
Every one of them is profitable or well-funded doing the opposite of what the incumbents optimize. What none of them has done yet is put it all in one place, on shared infrastructure, at platform scale. That gap — the openings nobody has fully built — is the next part.
Next in the series: Part 4 — three openings nobody has fully built, where the pieces above get combined into products that could actually scale. See all parts → · Get new parts by email
References
- Locket: App Store; funding, TechCrunch
- Retro: $21M Series A (TechCrunch); what it is and pricing
- Halide Process Zero: Lux Camera; TechCrunch
- Fujifilm Instax: Instax sales to exceed $1 billion (Digital Camera World)
Comments