An argument from
interest marketing.
The platforms settled the question of what distribution runs on. Then they kept the answer.
Exclusions first. This is an argument rather than a study. Every figure in it belongs to somebody else and is attributed where it appears and listed again at the foot. One of the two studies is published by a company we compete with. And the argument does not claim that owning an audience solves the problem of finding one. That part is dealt with plainly further down rather than left for a reader to catch.
Start with the measurements
Before any interpretation. Two reports published this year. Neither was commissioned by us.
receive fewer than a thousand views per post
receive fewer than a thousand views per post
receive fewer than a thousand views per post
name unreliable or declining social reach as a major strategic concern
The three reach figures come from the 2026 Creator Economy Report by The Influencer Marketing Factory with HypeAuditor and rest on more than five million creator accounts. YouTube Shorts is the healthiest surface in that study and 39.94% of its creators are still under a thousand views a post. The strategic-concern figure is from Circle’s 2026 Community Trends Report. Circle sells community software and competes with us. We quote them because the number is theirs rather than in spite of it.
The feed stopped asking who you know
This is not our observation and we are not going to dress it up as one. The industry has had a name for it for years. Ranking by the social graph asks who you are connected to. Ranking by the interest graph asks what you have shown an interest in. TikTok built a feed on the second around 2020 and the rest of the category followed it.
The honest version is narrower than the slogan. The interest graph did not replace the social graph. The two run alongside each other and any large platform uses both. What changed is which one decides whether a post is seen. That is enough to explain the four measurements above without reaching for a story about the death of social media.
A following used to be a distribution guarantee. It is now a weak signal that competes with every other signal.
The hinge
If distribution is decided by interest then interest is the unit of value. That is the part the platforms understood first and built for.
What follows from it is the part they had no reason to tell anybody. Every brand currently rents its interest. You bid for a segment. The platform keeps the member. The segment is re-let to a competitor the following quarter and nothing you paid for last year makes this year cheaper. Spend buys attention on a lease and the lease never converts to ownership however long you hold it.
So the claim here is not that social is broken. Everybody says that and it persuades nobody. The claim is narrower and harder to argue with: the platforms were right about interest and wrong about who should own it.
What an owned vertical does that a rented segment cannot
Not better. Different in kind. These five are structural rather than a matter of execution.
- It remembers. A feed forgets inside forty-eight hours. A vertical accumulates. The question somebody answered well in March is still answering in November.
- It has consequence. Members do things and are seen to have done them. No algorithm tracks whether you kept your word and no follower count records that you helped.
- It confers standing. Followers belong to the platform. Reputation earned inside a vertical belongs to the member and travels with them.
- It matches. Member to member. Work and hiring and deals and introductions. A feed can only broadcast and a vertical can introduce.
- It is addressable without permission. Email and notification to a member who asked for both on a domain nobody can deprioritise overnight.
The part of this argument that does not work
Owning a community answers retention and monetisation and data. It does not answer acquisition.
Circle’s own 2026 report finds that 67% of communities still meet their new members through social platforms. That figure has fallen from 76% the year before which suggests a direction of travel rather than a destination and one year is not a trend. Either way the honest position is the unglamorous one. Social becomes the top of the funnel rather than the whole of it. A brand that wants reach this quarter should buy advertising and we will say so rather than sell around it.
What changes is where the funnel ends. It used to end on somebody else’s platform. It can now end somewhere the brand owns and the difference compounds for as long as the community is worth belonging to.
The failure mode is not the software. It is a room with nobody talking in it.
What it asks of a brand
More than a purchase. A community is not a channel that can be switched on and measured at the end of the quarter. It is a room that has to be worth entering on a Tuesday when nothing is being sold.
That is the whole risk and it is worth naming rather than discovering. Most brand communities do not fail because the platform was wrong. They fail because the founding members were never recruited by name and nobody was responsible for the room being alive on the days when nothing was happening. Software does not fix that and content does not fix it either.
Which is the argument for treating it as a managed undertaking rather than a licence. The technology is settled and it is ours. The harder half is the part that keeps a room worth returning to and that is work rather than a feature.
Where this leads
Interest marketing is the category. A community on a brand’s own domain carrying the brand’s own name is the form it takes. Three of those are live today and every one of them is a click away rather than a case study.
Sources
- 2026 Creator Economy Report — The Influencer Marketing Factory with HypeAuditor, drawn from more than five million creator accounts The three reach figures and the YouTube Shorts figure quoted in the text.
- 2026 Community Trends Report — Circle The strategic-concern figure and the share of communities whose new members arrive from social platforms. Circle sells community software and competes with us.
- Interest graph and social graph — The Shelf Background on interest-led ranking and on why it sits alongside the social graph rather than replacing it.
All three read on 17 September 2026. Where a figure here disagrees with a later edition of the same report the later edition is right and this page is out of date.