Ads vs bidding in indie hacking

In 2026 the indie internet uses “ads” and “bidding” for two local products, and people also mean a third, older practice. Keeping them apart is the page. All three are ways to pay to be seen. The thing you are buying is different, and so is the thing the seller owns.

The state of the year is that shipping is cheap, so attention is what still has a price. These are the price tags.

Ads: a slot on attention that already exists

A classic ad is a publisher with an audience, selling a rectangle. The audience is the asset. The ad is a rental. Indie Hackers has sold sponsorships because the site has readers. A newsletter with a real list can charge for a mention. A jobs board can charge an employer to be seen by people already looking for work. Remote OK’s history sits in that older pattern: the product is the crowd, the fee is access to it.

TrustMRR added a twist. The page is a directory of revenue, often Stripe-connected, so the audience is people who came to look at numbers. Lou has written that sidebar ads made $20,378 within about three days of the late-October 2025 launch rush. That is founder-reported, from his newsletter, and it is a spike next to a launch, not a steady CPM you can paste into a plan. He also hoped a fee on acquisitions — he has said 3% — could replace ads. In the March 2026 recap, TrustMRR led his month at $36,000, which he tied to that fee as well as the product’s attention. Ads and a transaction fee can share a site. Verification did not abolish sponsorship. It made the sponsorship more valuable for a week, because the room was full.

The test for this kind of ad: if the publisher disappeared, would anyone still have a reason to look at that URL? If yes, the ad is renting a reason. If no, you are not buying an ad. You are funding a page that exists to hold the ad.

Bidding: the rank is the product

outbid.lol launched on 19 August 2026. Jonathan Wilke’s rule is the whole product. Pay more than the person above you and your link is number one. No revenue proof. No API key. A week-one recap, as written up by Steal What Works, put visitors around 1.36 million and bid revenue around $221,000, with hundreds of clones. Those numbers are reported, not a filing. afford.bid listing hundreds of boards by late August is the clone count as a public tracker, and trackers move.

This is not a publisher renting surplus attention. The auction is why people look. Spectators and bidders are the same crowd. When the spectacle fades, the “audience” fades with it. That can still be a lot of money in week one. It is a weak model for a product that needs the same buyers next quarter, unless the bid buys something that outlives the leaderboard — a joke, a brand impression, a backlink people argued was the point. Many bidders are other people running boards. The customer is the scene.

The ancestors are older than the costume. Alex Tew’s Million Dollar Homepage in 2005 sold pixels for $1. Marc Köhlbrugge’s highscore.money in 2016 let you pay any amount and wear that amount as a rank. 2026 did not invent pay-to-rank. It made a copy cheap enough to spawn a crowd of copies in a fortnight. Saturation in this corner is obvious because the clones are the content.

Lou’s HYROX auction is bidding that left the browser. Fifteen temporary tattoo slots, a doubling price, refunds minus fees if you were outbid, a closed total of $112,000 on hyrox.marclou.com. He named outbid.lol as the parent. The story has the brands and the rules. It is an ad in the sense that a logo sat on a body at a race, and a bid in the sense that the price was a contest. Hybrid products do not break the split. They show that the inventory can be skin, a laptop lid, or a fridge door. LivingAdSpace and the sticker variants were already turning objects into checkouts. The state is: if attention is scarce, sell the surface people are already staring at, and let the price be public so the price is also the post.

Performance ads: buying a visit

Google, Meta, and their cousins are the third meaning. You pay for a click or an impression in front of a query or an interest. Bootstrappers have always done this when the unit economics work. It is ordinary customer acquisition. It does not need a leaderboard, and it does not make you a publisher.

It is also the practice indie hackers complain about when they say ads got expensive and a small product cannot buy its way out. An Ask HN in February 2025, cited on the dead page, puts expensive ads next to tired SEO as the mood. That complaint is about this third meaning. It is not a review of outbid.lol, which did not exist yet.

Performance ads require a product that makes more than the click costs. Bidding among founders does not. A $4,000 bid to sit on a joke leaderboard can be rational as entertainment or as a campaign and irrational as “CAC.” Do not put both in one spreadsheet row labeled ads.

Week-one bid revenue is not MRR. A sidebar spike is not a yearly run rate. See MRR.

Which one the scene is in

The homepage calls the live plot bidding and ads because both are visible and both are about paying to sit at the top of something. The sharper 2026 novelty is bidding. Sponsor ads are old. Performance ads are old. A self-serve rank you can clone in a weekend is the chapter that opened in August.

For a product with a real buyer outside that chapter, the order of operations did not change. A direct path to the buyer beats renting a spectacle. An audience you already have beats paying for a look at other people’s audiences. Paying is rational when the look converts, or when the look is the joke you are selling. Getting paid by running the auction is a different seat from paying to enter it. Wilke was in the first seat for a week. Most of the clones were hoping to be. Most founders are in neither seat. They are trying to get a customer. The auction is the backdrop, not the assignment.

FAQ