Comparison was always the product

A fake private jet costs $34.99 an hour. A real one, grounded, going nowhere, costs $425.

Both prices are published. The first is FD Photo Studio’s weekday rate for Olympic 4, a plywood fuselage in Boyle Heights with artificial window lights, three flash units and a make-up station. The second is the twelve-hour rate for an actual Gulfstream III in Los Angeles, rented for photo and video shoots; shorter bookings run up to $750 an hour. The gap between those numbers is where the last decade of aspirational imagery was manufactured.

Nobody hid this. There is no leak, no whistleblower. The fake jet has a booking page. Peerspace lists jet interiors in Orlando, Yonkers, Atlanta and Levallois-Perret outside Paris, where the listing is titled, without embarrassment, “Studio Influencer | Intérieur d’avion Jet Privé”. Vivrelle rents rotating access to Chanel and Prada handbags for $45 to $309 a month, and closed a $62 million Series C in June 2025. Rent the Runway booked $329.8 million in FY2025 revenue across 143,796 subscribers. Lion Heart Lifestyle in Los Angeles runs a page headed “Models & Influencers” offering Lamborghinis, Ferraris, private jets “for aerial shots” and yachts — the only thing it withholds is the price list.

This is not a conspiracy. It is a supply chain, and it behaves exactly as a supply chain does when demand appears.

Every link is a rational actor

Follow the incentives and there is no villain anywhere, only people responding sensibly to what they are paid for.

A brand wants reach it can evidence, so it buys on follower counts and engagement rates. An influencer’s income therefore depends on numbers, not outcomes. Victoria O’Meara documented what happens next in Social Media + Society in November 2019, from 16 interviews with Instagram influencers — 15 women and one man, aged 24 to 41, with roughly 1,000 to 200,000 followers. They join engagement pods: private groups that coordinate likes and comments within the first five minutes of a post to game the visibility algorithm. One participant, “Sarah”, explained the logic without apology: “Brands are so numbers focused that an algorithm change… can have a dramatic affect on… income.”

If the numbers are the product, someone will manufacture the numbers. In Hanoi in 2023 the British photographer Jack Latham spent a month inside five click farms for his 2024 book Beggar’s Honey. He found two architectures: rooms of manually operated handsets, and “box farms” — phones stripped of screens and batteries, wired together and driven from one computer, so a single operator does the work of roughly ten thousand people. “They all looked like Silicon Valley startups,” he said. Services were advertised at under a cent per click, view or interaction. A test post he pushed through a farm drew over 6,600 likes against his normal baseline of 50 to 200.

That market has been priced in public for years. The New York Times‘ “The Follower Factory”, published on 27 January 2018 by Nicholas Confessore, Gabriel J.X. Dance, Richard Harris and Mark Hansen, found followers selling for about a cent each, and “high-quality” English-language bots with profile photos at $17 per thousand. Devumi had more than 200,000 customers and supplied more than 200 million Twitter followers. The FTC settled with Devumi and its chief executive for $2.5 million in October 2019, finding roughly 3.5 million automated accounts and over 58,000 orders for fake Twitter followers.

The photographs work the same way. A veteran Los Angeles street-style photographer told Bustle that event coverage is arranged in advance: “those shots don’t exist unless they’re sort of planned.” Cesar Peña of 247PapsTV described coordinating with a celebrity’s team and said of Kim Kardashian, “candid shots of her are very rare.” Tyler Williams, chief executive of Nouveau Communications, named the arbitrage: he had paid “a star $50,000 for a couple shots,” against a sponsored Instagram post from the same tier of talent costing “well over $1 million.” Same face, same product, one-twentieth the price, and the picture looks like it happened by accident.

The strongest objection, answered

The objection runs like this: none of this is new. Court painters slimmed their monarchs. Studio portraits were retouched with a pencil on the negative. Status display is a human constant, and treating it as a crisis is a moral panic with better graphics.

Most of that is correct, and it is why the outraged version of this argument fails. But three things have changed in kind, not degree.

The first is price. A court portrait cost a fortune, so the signal was honest even when the image was not — only a king could afford to be painted like a king. When the fake costs $34.99 an hour, the cost of appearing wealthy falls faster than the cost of being wealthy, and appearance stops carrying information.

The second is volume. A painting hung in one hall. A feed delivers a continuous, personalised, algorithmically ranked stream of other people’s best moments to a person lying in bed. The comparison set is no longer your village. It is a global tournament in which you are always in the bottom half, because the bottom half never gets ranked.

The third is detectability, and it is decisive. In 2018 Kleemans, Daalmans, Carbaat and Anschütz showed 144 Dutch girls aged 14 to 18 either original or body-reshaped Instagram selfies, in a randomised experiment published in Media Psychology. The participants could not reliably tell which photographs had been altered. They rated both sets realistic. They rated the manipulated ones higher. And their own body image was measurably lower afterwards. “Everyone knows it’s fake” is an empirical claim, and it was tested, and it is false.

Behind that sits the strongest causal evidence in the field: Braghieri, Levy and Makarin in the American Economic Review in 2022, exploiting Facebook’s staggered rollout across US colleges to show the platform’s arrival worsened student mental health, with unfavourable social comparison as the mechanism. The wider literature is genuinely contested and correlational effect sizes are small. That study is not correlational, and comparison is what it found.

Why “regulate it” is not the answer either

The instinctive response is law. It has been tried, and the record is discouraging.

France has required a “photographie retouchée” label on slimmed or widened commercial images since 1 October 2017, on penalty of €37,500 or 30 per cent of ad spend. Norway’s parliament voted 72–15 to require labels covering seven per cent of the image, in force July 2022 — then softened the guidance within two months under pressure from photographers. Israel has had disclosure rules for slimming edits since 2012. Between them, no meaningful enforcement action has been reported.

The pattern repeats in the United States. The FTC passed its Rule on the Use of Consumer Reviews and Testimonials on a unanimous 5–0 vote in August 2024, banning fake reviews, undisclosed insider endorsements and the sale of fake followers and views. As of this writing it has produced warning letters — ten unnamed companies in December 2025 — and no penalty judgment. The agency was careful to state the letters are not determinations of violation.

This is not cynicism about regulators. It is an observation about instrument design: rules requiring the inspection of billions of images for an invisible edit do not scale. The one intervention here with a documented outcome — the ASA banning two L’Oréal ads in 2011 after a single MP complained — worked because it put the evidential burden on the advertiser and offered a fast remedy. Everything else has been announcement.

What to do instead

Literacy, not outrage. Outrage assumes a villain and cannot find one; literacy assumes a market and reads its prices. Four habits, all cheap:

• Ask what the image cost to make. Not what it depicts — what it cost. Once you know a jet interior is $34.99 an hour and a handbag is $45 a month, a whole genre of photograph stops being evidence of anything.
• Check the ratio, not the total. Mae Karwowski of Obviously gave the working heuristic: “Say the person has 70,000 followers and they get 100 likes and zero comments per post — that red flag should go up.” Followers are the cheapest thing in this economy. Conversation is the expensive thing.
• Stop trying to spot the edit. The Dutch experiment settles this: you cannot, and confidence that you can is itself the vulnerability. Change the default assumption instead. Assume a commercial image of a body has been worked on, because it almost certainly has, and stop grading yourself against the output of a retouching pipeline.
• Run the ten-minute experiment on yourself. Hunt, Marx, Lipson and Young randomly assigned 143 undergraduates to ten minutes per platform per day for three weeks; the limited group’s loneliness and depression fell relative to controls. It is a small study with real limits. It is also a protocol you can replicate on a sample of one, for free, this month, with a phone setting.

Bow Wow got to the end of this argument before anyone else did. In 2017 he posted a photograph of a private jet and was debunked within hours by a passenger on his actual commercial flight. Eight years later, on Shannon Sharpe’s podcast, he said the plainest sentence in this entire archive: “I got caught, simple as that. I was fake flexing.”

The feed is a shop window. Everything in it is for hire, most of it by the hour, and the prices are on the internet. Comparison was never a side effect of the machine. Comparison was always the product. The only real defence is knowing what everything costs.