Positioning · Working Memo
EVERYONE Positioning
Why the integrated approach is the lane, who is in the broader landscape, and what we are downstream of.
- The inversion. We connect, we do not collect. Your data stays where it lives. The platform is a layer of intelligence and agency on top of Apple, LinkedIn, Gmail, Calendar, Facebook, and the rest. Trust pillars become structural facts of the architecture, not aspirational claims.
- Built for the agent era. As AI agents become primary users of digital systems, the platforms that win will be those whose data is most freely accessible to user-controlled agents. Walled gardens cannot match this without giving up their moats.
- Dual-layer composition. A private personal Roster delivers selfish benefit on day one, before any network effect. The public Team compounds value as members join. The composition is the strategic asset.
- The landscape is real, the integration is ours. Solid, MyData, DSNP, ATProto, Farcaster, Personal CRMs, Personal AI memory tools, agent-protocol layers, and civic deliberation tools all exist. No comparable effort has all five surfaces (film, book, campaigns, brand, network) composing as one. EVERYONE is the integration.
- We stand on a 25-year lineage. Identity Commons (2001), PGP Web of Trust (1992), MyData Movement, Solid (Berners-Lee), Plurality (Tang, Weyl). Acknowledging the lineage is what lets us be a serious participant rather than a naive newcomer.
Connect, do not collect.
Most platforms compete for your data. We invert that, structurally.
The dominant model wants your data on their servers, in their graph, with their lock-in as the moat. EVERYONE does not. Your contacts stay in Apple. Your professional graph stays in LinkedIn. Your messages stay in WhatsApp. Your calendar stays in Google. EVERYONE is a layer of intelligence and agency on top, never a competing repository. The architecture is the moat.
The Personal Roster v0.1 in production today reads from Apple Contacts (Apple still owns them), LinkedIn export (LinkedIn still owns it), Facebook export (Facebook still owns it), Calendar and Gmail signal (Google still owns them). The Roster stored in our Blob keyed by the user's auth is a derived view, not the source of truth. A user can stop using EVERYONE tomorrow and the network is still intact at every original source.
Five reasons this matters strategically
It dissolves the migration problem. Most platforms pitch "switch from X to us." We pitch "stay on X. We make X work for you." Zero migration cost equals zero adoption friction.
Trust pillars become structural facts. "Own your profile, no extraction, no exploitation" stop being marketing claims and become facts of the architecture. We cannot extract what we do not hold.
It sharpens the agent-era reframe. Walled gardens lock data in by design. EVERYONE connects across boundaries by design. We are the inverse of the moat.
It answers "what if I leave" by removing the question. There is no lock-in because there is nothing to lock. The lack of exit cost is itself the trust.
It defines the business model boundary cleanly. Revenue cannot come from owning user data because we never own user data. Revenue comes from access value, partnerships, marketplace activation, premium intelligence features. Aligned by construction.
It is reinforced by sliding-scale-as-default pricing across everything we make. The floor is what the system can sustain (manufacturing on physical, AI tokens on inference-backed features). The ceiling is what those who can give more choose to give. Free access exists when paid access funds it. The framing is "the team funds the team," not charity. Walled gardens cannot adopt this without breaking their advertising or surveillance business models. Subscription-by-MRR competitors cannot adopt it without breaking their investor narratives. The mechanism itself is part of the moat.
Built for the agent era.
AX is the new UX. We were designed for it.
Walled gardens are designed for human UX with intentionally restrictive APIs. That restriction is the moat. EVERYONE's architecture maps directly onto an agent-first world: trust pillars become protocol-level guarantees, the Roster has a public schema, agents can query freely under user consent. Walled gardens cannot match this without surrendering their moat.
LinkedIn obscures email by default. Facebook gives names and timestamps but no profile data via export. Even Apple Contacts is rich only because the user put data there themselves. The agent-readable layer across walled platforms is thin and fragmented, by design.
The agent layer is consolidating fast. Anthropic's Model Context Protocol (MCP) crossed 97 million monthly SDK downloads by February 2026 and was donated to the Linux Foundation in December 2025. Google's Agent-to-Agent protocol (A2A) was donated to the Linux Foundation in June 2025. The reference architecture (MCP for tool integration plus A2A for agent coordination) is now a pre-competitive standard. Walled gardens have to retrofit. EVERYONE was designed for this from the start.
What this looks like in practice:
- Your Roster is yours. Your agent (in any tool you use) can query it freely via clean MCP-style endpoints.
- The Connection Engine becomes more powerful in agent space. Your agent meets another user's agent, they negotiate connections at the agent layer with the trust pillars enforced as protocol rules. Multi-hop, agent-to-agent, with humans always being the destination.
- The pitch sharpens: the first network designed for the agent era. Your Roster is yours, your agent can use it freely, and the trust pillars hold whether a human or an agent is on the other side.
A private layer and a public layer, composing.
Personal CRMs are private and individual. Public networks are extractive. We are structurally both, in a way that does not collapse.
The private Roster gives the user selfish benefit on day one, before any network effect. The public Team compounds as members join. The composition is the strategic asset. Personal CRMs can become better individual tools, year after year, with a ceiling of "I have a great personal database." Public networks can become bigger, with the cost of extraction and migration. EVERYONE is structurally both, by design.
The private layer (Roster). Each user's intelligence layer over their own existing relationships. Connect-not-collect. Data stays at the source. AI is a broker, not a destination. The user gets value on day one before any other member joins.
The public layer (Team). Members opt into a public profile. Per-relationship visibility. The six-degrees graph. Asks-and-offers marketplace. The collective layer that compounds as users join.
The two compose. Paths between members work on the public graph with consent at every hop. Asks and offers live in the public layer. AI in the private layer can suggest "you should publish this ask, here's a draft." The Constellation visualization renders the personal Roster privately and the public Team publicly (where consent allows).
The competitive landscape, named honestly.
A view of who else is in this space, what each does well, and where the structural difference shows up.
The landscape is real and deep. Walled gardens, decentralized social protocols, personal CRMs, personal AI memory tools, agent-protocol layers, civic deliberation tools, and data-sovereignty pioneers all occupy adjacent territory. EVERYONE is differentiated not by being first to any single piece but by integrating across the layers in a way no other effort does, with cultural anchors (film, book, campaigns) that no infrastructure-only effort has.
Walled gardens (the incumbents)
Owns the data and gates the API. Business depends on this. Restricting agent access is the moat. Opening up agent access would compete with their own monetization. EVERYONE complements LinkedIn rather than replacing it: LinkedIn stays where the user's professional data lives; we read it and surface it intelligently in context across all the user's other sources.
Owns the social graph and the moat is the data. Messenger and WhatsApp depend on the user being inside their environment. EVERYONE reads what the user can export and surfaces it without competing for the storage. Meta cannot offer cross-source intelligence because that would require integrating with platforms they actively compete with.
Store the user's contact records but do not build intelligence on top. Apple's business is hardware, Google's is search and ads. A relationship-intelligence product does not move either revenue line meaningfully. EVERYONE adds the intelligence layer over what these stores already hold, without competing with them as storage.
Decentralized social protocols
The most mainstream-usable of the decentralized social protocols. Uses DNS domain names as handles, offers globally consistent engagement metrics and shared identity. EVERYONE could publish a member's public Team profile via ATProto without building our own social-graph protocol.
W3C standardized in 2026. The largest open social standard by adoption breadth. Threads added crossposting in 2024. Tumblr and Flipboard committed to support. Possible federation target for outbound publishing from EVERYONE.
Frank McCourt's decentralized social networking protocol, released 2021. A 20-million-member platform joined DSNP in 2025. If DSNP becomes the de facto open social graph protocol, EVERYONE's public Team should consider DSNP as a publishing target.
Each occupies a niche. Farcaster has identity onchain. Lens is Polygon-based. Nostr is relay-based, ~16M total users (~780K DAU), primarily Bitcoin community. All technically interesting; none currently positioned as our direct competitor.
Personal CRMs (the closest category)
The closest direct competitors as a standalone product. Each collects (you import, they store, the value lives in their database). EVERYONE connects (the value lives at the source). Personal CRMs are private individual tools. EVERYONE is dual-layer with a public Team that compounds. Personal CRMs never become a network. EVERYONE has the network as a structural layer. UpHabit (originally privacy-first personal CRM) pivoted to enterprise sales in 2022, a cautionary tale we design against structurally.
Yale-founded AI social network that lives in iMessage. User texts a phone number describing who they are and who they want to meet; AI returns a swipeable carousel of profiles. Explicit positioning: "the power of warm connections." Closest active competitor to the Roster + Connection Engine combination. Differences: their model requires others-already-on-platform from day one; our Roster works for one user with their existing data alone. Their substrate is iMessage; ours is the user's full multi-source data plus an opt-in public Team. Their positioning is consumer-social AI-native; ours is project-rooted in a film, a book, and a thesis; dual-layer in private Roster plus public Team; and agent-era by architecture.
Personal AI and memory tools
Adjacent but different focus. Rewind was acquired by Meta in December 2025 and sunset; Limitless continues as the surviving brand with a Pendant wearable. Tana raised $25M with a 160K+ waitlist building voice plus structured knowledge plus agents. None are relationship-graph products. EVERYONE is purpose-built for the relationship layer specifically.
"Human-centric frontier AI lab" with founders from xAI, Anthropic, DeepMind, OpenAI, Meta. Backed by SV Angel, NVIDIA, Bezos, GV, Emerson Collective. Stated mission: AI as "deeper connective tissue that strengthens organizations and communities." Closest thesis match in research-lab form to EVERYONE's "AI is a broker, not a destination" framing. Currently a research lab, not a consumer product. If they ship a consumer product, this is our closest peer-or-competitor. If they stay at the research layer, potential ally.
Agent-to-agent protocols (the substrate)
MCP standardizes how agents connect to tools, data sources, services. By February 2026: 97 million monthly SDK downloads, adopted by every major AI provider. A2A standardizes how agents discover, communicate, collaborate. The reference architecture (MCP plus A2A) is solidifying as a pre-competitive standard. EVERYONE's relationship-graph application is built natively on this stack; walled gardens have to retrofit, we are designing for it.
Civic deliberation
Audrey Tang and the g0v community built a consensus-finding system used in Taiwan. Of 26 national-issue deliberations on vTaiwan, 80% led to government action. Different problem space (consensus on policy questions among strangers vs intelligence-routing through known relationships) but adjacent mission. The Plurality Institute (Tang, Weyl) is intellectual peer territory.
Data-sovereignty infrastructure
Personal Data Pods. Real deployments in Flanders (entire region's digital economy), at a major US home-improvement retailer, in insurance. Open Data Institute took stewardship of Solid Project in October 2024. EVERYONE could run Roster storage on Solid pods as a future option, with Solid as plumbing rather than competitor.
The legal and standards scaffolding. MyData has 1,768 signatory organizations. EU eIDAS 2.0 mandates digital identity wallets by end of 2026. Decentralized Identifiers and Verifiable Credentials are W3C standards. EVERYONE uses these as plumbing for any cross-system identity claim.
The 25-year lineage we stand on.
We are not inventing the vision. We are operationalizing one that has been articulated for decades.
Identity Commons formed in 2001. The Internet Identity Workshop has run twice yearly since 2005. PGP Web of Trust was articulated by Phil Zimmermann in 1992. MyData has 1,768 signatory organizations. Solid raised $30M in 2021. EVERYONE is downstream of all of them. Acknowledging the lineage is what lets us be a serious participant rather than a naive newcomer.
Identity Commons (2001). Founded by Kaliya Young, Doc Searls, and Phil Windley to evangelize decentralized user-centric identity infrastructure when very few people were thinking about these issues.
Internet Identity Workshop (since 2005). Runs twice yearly, now at 37+ events. The user-centric identity community is older than most current Web3 efforts by a decade.
PGP Web of Trust (Phil Zimmermann, 1992). The original "trust as protocol, not platform" idea.
MyData Movement (NGO, growing since 2017). User-centric data governance framework. 1,768 signatory organizations. Strong policy presence in Europe, where the EU Data Act (effective September 2025) extended sovereignty beyond personal data to industrial and connected-device data.
Doc Searls and the VRM project. Vendor relationship management as the inverse of CRM, originating around 2006.
Solid (Tim Berners-Lee, Inrupt). Personal Data Pods. Inrupt raised $30M Series A in 2021. Open Data Institute took stewardship of Solid Project in October 2024.
Plurality (Audrey Tang, E. Glen Weyl, 2024). Book published openly on GitHub, CC0, translated into more than a dozen languages. The closest current articulation of plural democratic technology.
DWeb Camp (Brewster Kahle, Internet Archive). The community gathering where the lineage convenes in person. Camp 2026 runs July 8-12 in Berlin.
We inherit the vocabulary (user-centric, self-sovereign, data sovereignty, web of trust), the credibility of building on a tradition rather than claiming to invent, and the community of people who have been working on this since before most current investors were paying attention.
What we add: a film, a book, campaigns, brand, and a consumer product (the Roster) that surfaces the philosophy as immediate selfish benefit, rather than a B2B compliance pitch. None of the data-sovereignty efforts have a film. None of the decentralized social protocols have a thesis carried by book and campaigns. We are the application layer with cultural carriers.
The lane we hold.
What is genuinely ours is not any single piece. The pieces exist. The integration, anchored in cultural meaning, is what is new.
EVERYONE is the first attempt at a relationship-graph project where the trust pillars are protocol-level structural facts, the consumer surface delivers selfish benefit on day one, the network layer composes opt-in publicly, and the cultural framing comes from a film, a book, and a campaign-driven team-forming mechanic, not just a brand brief. The integration is the lane. The cultural anchor is the moat.
What no other effort has, integrated
A cultural artifact in production. A film going to theaters, a book in manuscript, campaigns including the Biggest Team that is currently being developed for the late 2026 launch window. Distribution channel none of the named competitors have.
A canonical thesis. "The truth that we are connected is structural." A frame that can be quoted and returned to. Most products have features. We have a frame.
The recognition mechanic. EVERY1 numbers and "the team forming" gives the network meaning before the network is dense. Series matches strangers via Tinder-style swiping. We hand someone a number and tell them they are part of something already true.
A working v0.1 over real data. Not a research lab. Not a pitch deck. Operational since May 10, 2026, with 11,518-person Rosters and live AI search.
Why incumbents cannot copy without giving up their moats
Every walled garden's business depends on owning the user's data. LinkedIn's professional graph. Facebook's social graph. WhatsApp's message corpus. Each is monetized either directly (advertising on the data) or indirectly (lock-in that drives engagement). To copy the connect-not-collect architecture, they would need to open APIs that let users access their own data freely from third-party tools, stop collecting and storing the user's network as a centralized graph, and compete with themselves by enabling cross-platform intelligence layers. This is a business model inversion, not a product change. They will not do it because their entire valuation depends on not doing it.
Walled gardens have scale and capital. EVERYONE has architecture and trust. The architecture is incompatible with the walled-garden business model. The trust is incompatible with the walled-garden track record. Capital and scale advantage compound when the field is comparable. They become disadvantages when the structural premise is the differentiator.
Allies, not just competitors
Most of the efforts named above are more ally than competitor. We can integrate with Solid Pods as a storage layer for the user's Roster export. We can publish a member's public Team profile via DSNP or ATProto. We can use DIDs and VCs for any cross-system identity claim. We can expose the Roster as an MCP server. We can join the Internet Identity Workshop community to be part of the conversation. We can attend DWeb Camp. The lineage is not opposition; it is the substrate we build on top of.