NOTE Secret — not listed in public feeds

Medium’s business model: A Level 3 Aggregator

This was originally published as an internal company blog post.

Medium’s business model: A Level 3 Aggregator

Author: Jacob Bennett
Date: December 11, 2025
Tags: Context sharing, Thinking out loud

I think the easiest place to start with Medium is understanding that we are a Level 3 Aggregator (defined in this Stratechery article — please read this). Aggregators are a specific type of platform with three distinct characteristics:

  1. Direct relationships with users. The platform has the relationship with the user vs the creator nurturing a relationship directly.
  2. Zero marginal cost for serving users. i.e. an internet company selling digital goods, where cost of goods sold (variable cost to serve a single user) is 0.
  3. Demand-driven multi-sided networks with decreasing acquisition costs. As more consumers (users) join the platform, suppliers join the platform on the platform’s terms and effectively commoditize themselves.

…suppliers can be commoditized leaving consumers/users as a first order priority. By extension, this means that the most important factor determining success is the user experience: the best distributors/aggregators/market-makers win by providing the best experience, which earns them the most consumers/users, which attracts the most suppliers, which enhances the user experience in a virtuous cycle. (source)

More simply: aggregators own the user relationship, and scaling that relationship is what brings value to the platform.

What does being an aggregator mean for Medium?

It’s important to understand we’re an aggregator because the problems we face are known and often already solved by other platforms.

As an aggregator, the product we sell is not the content (posts), it’s the bundle of content (i.e. the aggregation). We add value to the bundle by curating it and increasing the relevance of individual bundles to the readers who are buying it. I.e. the value we give to readers is superior curation and discovery (this is also where most other aggregators provide value to users).

Medium takes the paywall approach (vs ad-based revenue) to monetize the platform. We believe that ad revenue misaligns incentives and leads to a worse user experience (Vlad Prelovac (CEO, Kagi) wrote a great piece on this idea). We chose to build a platform with a single paywall for access to our full aggregated (and curated) content library.

Challenges and risks for Medium as an aggregator

Success as an aggregator is all about the relationship with the user.

In our case, the user is a Reader. The Reader Experience is the determining factor in Medium’s success. All product features and work streams should ultimately be in service of the Reader.

  • The Partner Program keeps the quality of content on the platform high to increase our ability to curate good content for readers.
  • The For You Feed helps users discover more good writing and move from a narrow focus to more broad reading habits.
  • Publications serve as topical curation for readers.

Highest risk areas for Medium:

  • We do not own our paywall. This gives up a lot of the highest-leverage control over the quality of the bundle we sell.
  • We are very slow at reacting to users and adding features that would improve the Reader Experience.
  • Our earned acquisition channels (SEO) are dead or dying, and our owned channels are not strong enough to sustain growth.
  • Our fixed costs are very high (~$0.07 per daily active user).

Opportunities for Medium as an aggregator

Aggregators enjoy winner-takes-all effects. As demand grows, the value of the platform to suppliers increases. Suppliers join (commoditized) and increase the value of the platform to users. We should see user retention increase over the long term as competitors find it very difficult to pull users away. We have a huge lead in the market (thank you, past Medians). We have time to solve problems (thank you, brief period of profitability). And if we can solve them, we will win and take it all.

Zero marginal user costs and zero marginal supply costs means we can effectively capture the full value of user growth (i.e. each additional paying member adds $4.35 to our bank) and fixed cost reduction. I’m pointing out a bit of the obvious here—we’ve said for years that cost savings and user growth have the same effect and should both be valued against our same impact measurement: ladder steps.

Comparison to other platforms

Other aggregators

Reddit

Is Reddit an aggregator?
User (reader/lurker/poster) relationship is with the platform primarily✅ Redditors have very high brand loyalty to Reddit. Creators on the platform are not differentiated (individuals can have karma, and badges within a community, but otherwise all users look identical). Readers read the content on the platform and don’t leave the platform if a creator leaves.
Zero marginal cost for serving users✅ Yes. Hosting costs are high because of the scale. But serving a marginal user is effectively free.
Demand-driven multi-sided networks✅ Creators post to reddit because that’s where users are. Creators karma-farm because users will engage with their content.

Reddit is a level 3 aggregator. And arguably our strongest competitor and threat to our company’s mission.

YouTube

Is YouTube an aggregator?
User (viewer) relationship is with the platform primarily✅ Viewers may have high loyalty to creators. But viewers are loyal to the platform first and foremost.
Zero marginal cost for serving users✅ Effectively. There is no COGS for serving a user. Or the COGS tends to zero (with fixed costs high). YouTube hosting costs are astronomical (storing, parsing, caching video).
Demand-driven multi-sided networks✅ Creators come to YouTube because that’s where the viewers are. Individual creators have effectively zero power, they are there on YouTube’s terms.

YouTube is also a level 3 aggregator. Their business is primarily ad-based. They pay creators in (1) reach and (2) ad revenue share.

They solved the quality problem with an amazing recs engine. They also prevent monetization for channels under 10k subscribers. (Medium used to require 100 followers.)

Spotify

Is Spotify an aggregator?
User (listener) relationship is with the platform primarily✅ Listeners may have high artist loyalty. But they also use the platform for all other artist discovery, podcasts, etc.
Zero marginal cost for serving users✅ Effectively. There is no COGS for serving a user. Or the COGS tends to zero (with fixed costs high.

*Probably something to talk about wrt their rev share model and how to think through COGS. | | Demand-driven multi-sided networks | ✅ Artists syndicate to Spotify because that’s where listeners are. Artists and labels usually join on Spotify’s terms. Larger labels can negotiate better deals. |

Spotify is a level 3 aggregator. Their business has an ad-supported free tier, but their largest source of revenue is their paid tier.

They solved the quality problem with an amazing recs engine. They also don’t rely on discovery as much to retain listeners. Artist fans are pretty sticky.

NY Times

NY Times is a level 1 aggregator (they own their supply, but they also own the reader relationship). They use a hybrid subscription + advertising revenue model.

Multi-sided marketplaces

Substack

We compete with Substack for supply. But the product we sell is different (we actually sell a product… Substack is a platform where writers sell their own products (posts)).

Etsy

Classic two-sided marketplace. We don’t match this at all, and I’ve left this here to emphasize how much Medium is not a two-sided marketplace.

Bandcamp

Left here to highlight the difference between an aggregator (Spotify) and a multi-sided marketplace (Bandcamp) in the same industry (music). Where Spotify sells a bundled subscription to a platform, Bandcamp lets artists sell directly to fans. Bandcamp doubles down on the artist-fan relationship because that’s what a multi-sided marketplace is good at. Spotify doubles-down on the listener’s experience on the platform because that’s what an aggregator is good at.