PESO Model: Paid, Owned, Earned Media Against Platform Risk

Why the PESO model is a risk model in 2026

The PESO model comes from PR and sorts channels by ownership: paid (bought), earned (won), shared (distributed on social platforms), owned (your own). For a long time it was read as a communications plan: which message runs on which surface. That reading falls short. The real question the model answers is: who owns access to your audience?

On social platforms the answer is: not you. Followers are not a possession, they are access granted until revoked. The feed is rented ground, and the landlord changes the terms of the lease whenever it suits. At the end of 2024 LinkedIn rolled out 360Brew, a foundation model that matches posts to interests semantically instead of following the follower graph. Instagram distributes feed, Reels, Stories and Explore through four separate ranking systems that sort by predicted interest. In January 2026 TikTok started retraining its recommendation algorithm after the Oracle deal for the US business. The pattern is the same everywhere: every post competes again from scratch, and the rules change without notice.

What that means in practice is on display at XING. The platform scrapped groups and events and wound down the newsfeed. Companies that had built a community there over years were left without a distribution channel. Not because of an algorithm update, but because of a business decision by the operator. That is exactly the class of risk the PESO model addresses once you read it as a hedge.

The four channel roles, realistically distributed

The classic PESO diagram treats all four fields as equals. In 2026 that no longer holds. Organic reach on company pages has structurally eroded, and the benchmarks for that are in the article on organic reach and pay-to-play. The roles have shifted accordingly.

Shared: Organic social presence is the laboratory. This is where you test which topics and formats resonate before you put budget behind them. And this is where trust is built, above all through personal profiles. On LinkedIn, profiles clearly beat the company page, which is why corporate influencer programmes are the strongest organic lever; how to set up such a programme is described in the article on corporate influencers and employee advocacy. But even a strong profile belongs to the network, not to you.

Paid: Reach for new audiences is a budget question in 2026. Anyone planning without a media budget is planning the loss of reach along with it. Paid is the engine that creates first contacts and fills the owned list. The dependency remains all the same: Meta's ad systems Andromeda (retrieval) and GEM (ranking) reward broad targeting and punish frequent edits with a reset of the learning phase. You buy reach on terms the seller sets.

Owned: Website, newsletter and email list are the only channels whose reach no platform operator rewires overnight. A send goes to the entire list, minus bounces and spam filters, regardless of what a feed algorithm favours this week. That is the core of the hedge argument: owned is the only quadrant in which you actually own reach.

Earned: Press, shares, recommendations and, increasingly, mentions in AI answers. Earned cannot be planned, but it is the quadrant with the highest credibility per contact, because the message comes from a third party.

Owned media as a counterweight: newsletter, email, community

Most companies in the DACH region treat owned media as an obligation. A newsletter exists because it has always existed, the list grows through a footer form, and nobody measures how many social contacts ever made it onto that list. That is the gap the PESO model makes visible.

Building it follows a simple logic: every social content series needs a route into an owned channel. That can be a newsletter, a download with email opt-in, a webinar registration, a community on a platform such as Discord or Circle, or a WhatsApp broadcast. Which owned channel fits depends on the audience. For B2B decision-makers, email is the most reliable route. For consumers, messaging is gaining weight: WhatsApp Channels and broadcasts are an underrated channel close to first-party data, in our assessment one of the more substantial trends for 2026 and 2027.

Three rules for building owned:

• Treat the opt-in as a conversion goal: A newsletter sign-up is a lead, even when there is no purchase intent behind it yet. Measure it per social channel and per content format, and you will see which content turns reach into possession.
• Keep owned content independent: A newsletter that only recycles social posts gives nobody a reason to prefer the list over the platform. The owned channel needs content that appears there first or only there.
• Secure deliverability and legal basis: Owned channels have dependencies of their own (spam filters, double opt-in, GDPR). They are smaller than the platform risk, but not zero.

Community platforms are substantial for niches and membership models, but they are not a reach channel for the mass market. The Fediverse (Mastodon, Bluesky) works as a resilience hedge against platform dependency, but for most B2B brands it does not yet deliver reliable reach. Both are observation posts, not a replacement for the email list.

Earned media, dark social and AI answers

Earned media has two faces in 2026. The visible one: press, backlinks, public shares. The invisible one: dark social. Most sharing happens in direct messages and WhatsApp groups and turns up in no analytics tool as a social source. At the same time, sends and shares on Instagram and TikTok have become the leading signal; according to Adam Mosseri, sends per reach count for considerably more on Instagram than likes when it comes to reach beyond your own followers. Dark social is therefore blind spot and ranking driver at once. Sendable content wins twice, you just cannot see it in the dashboard. As a rule of thumb: if the question “How did you hear about us?” names social for more than 20 per cent of leads while last click shows under 5 per cent, the dark social effect is confirmed. How to combine self-reported attribution, MMM and incrementality for this is set out in the pillar on social media analytics and measurement.

The second shift concerns where earned media takes effect. According to the Reuters Institute Digital News Report 2025, 7 per cent of respondents internationally use AI chatbots as a news source weekly, and among the under-25s it is 15 per cent. That is still small, but it grows upwards through the age cohorts. Earned media therefore increasingly means appearing in the answers of ChatGPT, Perplexity and Google AI Overviews. How strongly mentions on third-party sites carry through there is covered in the existing article on earned media and AI citations. The SEO mechanics behind it, meaning citability, source authority and zero-click logic, are in the article on zero-click strategy.

For the PESO model, the consequence Rand Fishkin summed up in a blog title is enough: In a Zero-Click World, Traffic is a Terrible Goal. His data basis is US clickstream; in our assessment the mechanics apply in the DACH region in exactly the same way. When platforms and AI systems pass the click through to your website less and less often, website traffic is unusable as a measure of success for social and earned. The goal shifts to presence where the audience already is, and to converting that presence into owned contacts.

Scenarios for a loss of reach

A pre-mortem per channel is the operational core of the PESO hedge. The question is not whether a loss of reach will come, but which one comes first. Three scenarios cover most cases.

The first scenario is the most common. The claim that an update can halve reach overnight is experience, not measurement, but the direction is right: interest graph systems distribute by predicted interest, and a new model predicts interest differently. For planning that means organic reach is a variable with a wide range and owned reach is a constant. Budget and targets should reflect that.

Most people underestimate the second scenario until it happens. A suspended company page with 20,000 followers is a total loss if none of those contacts is on the email list. The scenario forces an uncomfortable calculation: how many of your followers could you reach tomorrow without the platform? If only a fraction of them also exists as an owned contact, the PESO balance is off, however good the organic numbers look.

The third scenario is no longer a thought experiment in the DACH region. XING demonstrated it. The early indicators were visible: functions disappeared, the positioning shifted towards recruiting. Anyone who systematically moved community contacts into owned channels back then lost little.

Typical mistakes in assigning the roles

The mistakes repeat across industries and company sizes, and the most expensive one is always the same: owned runs as a by-product instead of being the goal of the social work. A lopsided channel mix or a media budget that is too small can be corrected within a quarter, a missing email list cannot. Build it only once reach has already collapsed and you are building it without reach.

• Owned as an afterthought: The newsletter and the list are maintained, but not measured as the goal of social activity. The hedge then stays theoretical.
• Treating social as owned: Follower counts are reported as possession although the platform controls every contact. Vanity metric and risk blindness in one.
• Paid as a substitute for owned: Budget compensates for organic losses until the CPMs rise. Without building owned, you buy the same audience again every quarter.
• Understanding earned as press only: Dark social and AI citations are missing from the model, although a growing share of referrals happens there.
• No scenarios: The strategy assumes stable reach. Every algorithm change then looks like an accident instead of a case you planned for.

Measurement: how to check the balance

The PESO model needs no new tools, but it does need three metrics that are missing from most social reports.

Owned conversion rate: The share of social reach that ends in newsletter sign-ups, community joins or broadcast opt-ins, per channel and per format. This is the metric that quantifies the hedge.

Reachability without the platform: Owned contacts in relation to the follower base. If it rises, platform risk falls. If it falls despite follower growth, dependency is growing.

Self-reported versus last click: The gap between what leads say about their source and what the tracking shows. It makes dark social visible and shows whether earned and shared deliver more than the dashboard reports.

These three numbers belong in the quarterly strategy review, not in weekly content reporting. They change slowly and answer a strategic question: do you own your reach, or are you renting it?

Conclusion

In 2026 the PESO model is less a communications plan than a balance sheet. On one side sit shared and paid, where reach is created but never belongs to you. On the other side owned, where reach becomes possession, and earned, where it gains credibility. Work only the first side and you have a strategy that gets rewritten with the next model update. Run social consistently as a feeder for owned channels and a halving of reach costs you impressions, but not the audience.