Most Companies Are Wasting Their Stack
I started my presentation this year by playing an AI-generated song I made with exactly two credits. It was terrible, but it perfectly summed up the...
4 min read
Writing Team
:
Jan 26, 2026 8:00:00 AM
For twenty years, the internet centralized. Facebook, Twitter, Instagram—a handful of platforms captured billions of users. Marketing adapted by mastering these concentrated attention pools. Buy ads on Facebook. Optimize for Instagram. Tweet strategically. Simple.
That era is ending. The web is federating again.
Federated systems distribute control across many independent servers that communicate through shared protocols. Email is federated—Gmail and Outlook users can message each other despite using different providers. No single company controls email.
Mastodon, Bluesky, and emerging ActivityPub-based platforms apply this model to social media. Instead of one Twitter with 500 million users, you get thousands of smaller servers (instances) with various communities, all able to interact through common protocols. No company controls the entire network.
This sounds democratic and healthy. For marketers, it's a nightmare. You can't "advertise on Mastodon" the way you advertise on Facebook—there's no central ad platform, no unified analytics, no single algorithm to optimize for. Federation fragments everything centralized platforms made scalable.
Platform marketing worked because platforms were monopolies. Everyone was on Facebook, so brands went there. Twitter had the conversation, so PR teams monitored it. Instagram owned visual culture, so brands created there.
Federation destroys monopoly advantages. When users can choose between hundreds of instances running similar protocols, no single platform captures everyone. Community contribution versus promotional content becomes crucial—you can't buy your way into federated spaces the way you bought Facebook ads.
Bluesky's growth demonstrates this shift. It's not trying to become the next Twitter monopoly—it's building open protocols anyone can build on. If successful, this means "marketing on Bluesky" won't mean what "marketing on Twitter" meant. There won't be one Bluesky to target. There will be hundreds of Bluesky-protocol instances, each with different cultures and rules.
Federated marketing means understanding instance cultures individually. Mastodon instances range from tech-focused to arts communities to geographic regions to special interests. Each has distinct norms, moderators, and tolerance for promotional content.
What works on one instance gets you banned on another. There's no universal strategy. You're back to relationship-based marketing, understanding specific communities before participating. This doesn't scale the way traditional versus modern go-to-market approaches do on centralized platforms.
Some instances explicitly ban brands and commercial activity. Others welcome thoughtful brand participation. You need to research each community separately, the way marketers once researched individual forums before Reddit consolidated forum culture into subreddits.
Federation's strength is content portability—users own their data and can move between instances. Marketing's challenge is that content doesn't travel with platform-specific optimizations. Your viral Twitter thread doesn't port cleanly to Mastodon's different culture and character limits. Instagram aesthetics don't translate to federated image platforms with different discovery mechanisms.
You can't optimize once and distribute everywhere. Each federated platform requires native content creation that respects its specific culture. Social media content strategies built for centralized platforms assume you can create once, optimize for algorithmic distribution, then reap scalable returns. Federation breaks all those assumptions.
Centralized platforms have algorithms you can study and optimize for. Federated platforms have... whatever each instance administrator decides to implement. Some use chronological feeds. Others implement various discovery mechanisms. Some have no algorithm at all—just pure chronological timelines.
Visibility engineering becomes impossible when there's no consistent algorithm to engineer around. You're back to earning attention through relevance and relationships rather than gaming platform mechanics.
This sounds idealistic until you realize it means marketing on federated platforms requires actual value creation. You can't shortcut to visibility through paid placement or algorithmic manipulation when communities control their own spaces and can defederate (block) entire instances engaging in spam or manipulation.
Many users flee to federated platforms specifically to escape surveillance capitalism. They don't want targeted ads, don't want behavioral tracking, don't want their data monetized. These users chose federation partly to avoid marketing.
Privacy-first messaging apps showed this tension already—marketing to users who explicitly opted out of marketing surveillance. Federated platforms amplify this challenge because community norms often explicitly reject commercial activity.
Your marketing presence requires explicit permission and value contribution, not just advertising spend. This fundamentally changes ROI calculations when audience building requires months of authentic participation rather than weeks of ad optimization.
On centralized platforms, your account exists at platform discretion. Twitter can ban you. Instagram can delete your page. Facebook can restrict your reach. Federation inverts this—you can run your own instance, control your brand's infrastructure, own your audience relationships.
Some brands will run branded instances—company-controlled servers federated with the broader network. This requires technical capability and ongoing maintenance but provides genuine ownership. You're not renting attention from platforms anymore.
The trade-off: running your own infrastructure versus paying platforms for infrastructure and attention. For brands with technical resources, federation offers genuine digital ownership. For brands without those resources, it's another complexity multiplying an already complex marketing landscape.
Federation fragments audiences intentionally. This protects users from surveillance and manipulation. It also makes marketing harder, which federation advocates consider a feature, not a bug. They want marketing to be harder when "easier marketing" meant invasive tracking and algorithmic manipulation.
The question isn't whether to embrace federation—it's whether your brand can thrive when the internet stops consolidating audiences into platforms you can buy access to. If your marketing depends on paid placement and algorithmic distribution, federation represents existential threat. If your marketing depends on genuine value creation and community relationships, federation might be your competitive advantage.
Centralized platform marketing is dying slowly. Federation accelerates this death by providing viable alternatives that explicitly reject centralization's marketing conveniences. Brands built on platform dependence face existential questions. Brands built on audience relationships face tactical adjustments.
The federated web isn't hypothetical future—it's happening now through Mastodon, Bluesky, and emerging ActivityPub adoption. Marketing strategies that assume platform consolidation won't survive audience distribution.
Want to build marketing approaches that work when audiences fragment across federated platforms? We help companies adapt to decentralization rather than depending on consolidation. Let's talk about marketing that survives distribution.
I started my presentation this year by playing an AI-generated song I made with exactly two credits. It was terrible, but it perfectly summed up the...
The best technology disappears. It becomes so seamlessly integrated into our environment that we stop noticing it's there at all.
Let's talk through marketing technology and understand how each piece fits into the puzzle of data-driven marketing success.