Compound or trend.

The orientation

Why we built this

Every privacy founder I talk to says some version of the same thing. The messaging isn't landing. They can't say why. Their advisors give vibes. Their investors say make it punchier. Their marketing teams hand back work that pulls a "try again, I'll know it when I see it." Nobody can name the problem because nobody is treating narrative as something measurable.

This is the approach that does. At F/AS we treat narrative as architecture, not art. To our knowledge, this is the first peer-calibrated read on narrative position and quality the privacy sector, or any digital-asset sector, has ever had.

The privacy sector has spent the last eighteen months building. New protocols launched. Old protocols matured. Token economies emerged. Stablecoins, mainnets, encrypted auctions, regulator-shaped framings, and Cardano-anchored launches all happened inside a window short enough to fit in one report.

In the same eighteen months, a US administration changed its posture toward digital assets, the EU operationalized "Going Dark" surveillance proposals, the UK passed an age-verification regime that pulled some projects into direct policy advocacy, and state-actor campaigns targeted Signal and WhatsApp accounts. At the same time, the institutional buyer who writes the largest checks for privacy infrastructure started to show up in earnest. The work the field did to communicate with those buyers, in the channels those buyers actually evaluate, is what this report covers.

I'm Devon Ferreira, Founder and Principal at F/AS, a boutique strategic advisory firm. We help early-stage teams in digital assets, AI, and infrastructure develop positioning, narrative, and go-to-market strategies that compound. I was the first CMO at Avalanche, where I built the brand, and CMO at Immutable, where I led a brand realignment through their zkEVM launch and studio consolidation. Over the last five years I've advised dozens of early-stage blockchain and AI teams.

The question this report is built around, what the field actually looks like when you measure how it tells its story, is the question founders ask me most. This time, I answered it with data.

What we measured

Two scores. Four stories.

There are two questions worth asking about any project's narrative work. Is the story landing, and is the story distinctive? Most positioning advice in this sector treats those as one question, which is how founders end up with messaging that's either polished and forgettable or sharp and incoherent. We treat them as two. We score each separately, calibrate each against the rest of the field, and put both on the signature chart that anchors this report.

The vertical axis

Narrative Strength

The signature score in this report is Narrative Strength, abbreviated NS. It's a 0-100 rating of how clearly each project tells its story across the surfaces its audiences actually use to evaluate it. Six dimensions go into the score, each weighted by editorial importance.

The score is calibrated against the field. Strong work lands at 80 and above. Solid foundation at 70 to 79. Mid at 60 to 69. Emerging below 60. Three projects in the field could not be scored at the corpus level for reasons we explain in the methodology section, and we substitute a website-only score for those three. Their substitution is editorial signal in itself, and we treat it that way.

The horizontal axis

Differentiation

Narrative Strength tells you whether the story lands. Differentiation tells you whether the story is yours. A project can tell its story cleanly and still sit close to the field's center of gravity, where its peers tell variations of the same story. A project can hold a position no peer holds and still not be expressing it cleanly. Differentiation is about the position, not the polish.

We compute Differentiation in three steps.

The result is a Differentiation score per project. High Differentiation doesn't mean better. It means distinctive. A project can be highly differentiated and still not land its story (we have a quadrant for that). A project can land its story cleanly and still sit close to peers (we have a quadrant for that too).

The two scores together place each project on the signature chart.

The conversation

What the privacy sector is talking about

The privacy sector is having twelve major conversations at the same time, with twenty-two sub-conversations branching off them. We didn't pre-load any of those conversations. The methodology extracted the taxonomy directly from the corpus and validated it against per-project coverage until every cohort project mapped to at least one active theme. What you see when you ask the data what the field talks about is its own conversation, in its own language.

Narrative trajectories

Where the wind is blowing

A year ago the privacy sector was defined by what it stood for. Today it's defined by what it's building, with the principles still acting as the foundation underneath. The shift didn't happen because the field changed its mind about its values; it happened because the conditions finally caught up with them. The institutional buyer who writes the largest checks for privacy infrastructure started showing up in earnest. The regulatory ceiling that hung over privacy coins for years began to lift. The technology that was theoretical for a long time became something you could actually ship. The price action returned to validate the thesis the field had been holding through the bear. What you see now when you scan the field's twelve loudest conversations is a sector that has stopped having to justify its existence and started shipping what the existence was for.

Where we see the strategic divergence

The most striking patterns surface where the field disagrees about what privacy is for. The methodology surfaces those disagreements as sub-theme emphasis pairs at high surprise score within shared root themes. Three divergences carry the report.

What the data says

The privacy sector is moving toward programmable privacy with institutional infrastructure as the commercial foothold. It's divided about whether the technology serves compliance or composability. It's winning audiences but mostly not yet pulling proportional trade press for the work the projects are doing.

This is the reference point. Differentiation, the horizontal axis on the signature chart, is each project's distance from this field-level conversation. A project sitting close to the center of the conversation lands at low Differentiation. A project sitting at an angle to it, by choice or by accident, lands at high Differentiation. The four-quadrant placements emerge from this work.

The findings

Three findings worth keeping

Three findings carry the report. Each one was a surprise we found in the data, sharp enough that it earned a full section. If you read nothing else, read these.

Finding 1

Attention without compounding is the sector's most expensive habit.

The privacy sector is winning the velocity channel. Most of it hasn't yet built foundation narrative on the surfaces that compound across audiences and across cycles.

The numbers behind that claim are straightforward. Average sentiment on X across the cohort lands at just under four out of five, which is to say the audience is leaning in. The typical project carries about two-thirds of the narrative beats strongly on its X feed, and thirteen of the fifteen corpus-scored projects post Narrative Strength of 65 or higher on that surface. Aleo NS 90+ on X. Midnight NS 85+. Aztec and DuckDuckGo NS 75+. The field isn't phoning anything in.

The website lands at about half that strength. The typical project carries roughly a third of the narrative beats strongly on the homepage. The buyer who arrives at the homepage from a viral thread finds a story half as deep as the thread that brought them there.

What the field has won, and what it hasn't

Three observations worth holding in mind before we get to the gap.

The X-feed discipline is real. All fifteen corpus-scored projects post NS 50 or higher on the velocity channel, and thirteen of them above 65. The voice is sharp, the cadence is consistent, and the audience is leaning in.

Reception is favorable. Average sentiment is just under four out of five. The field is winning audiences over rather than annoying them.

Volume is structurally appropriate to the medium. X is temporal by design; websites are written once and revised over time. Volume isn't the asymmetry. The asymmetry is in the quality of narrative carry per surface.

That's the frame. The field has the talent. The reception is favorable. What most projects don't yet have is anything to point the favorable reception at.

Two projects break the pattern

Mullvad VPN. Website Narrative Strength runs at parity with the X feed. Only project in the cohort where that's true. Mullvad's audience doesn't live on X. They live in search results, expert roundups, and the recommendation a friend makes when you ask which VPN you should actually use. Mullvad turned off Visa and Mastercard. They accept cash by mail. They optimized for the channel that closes them, even when the channel is friction itself.

Aztec. Aztec's website Narrative Strength sits well above the next-best peer in the programmable-privacy stack. Aztec isn't website-anchored the way Mullvad is. But the website is doing more narrative work than any other project in its peer set, by a wide margin.

Two projects, two completely different parts of the field, two completely different audiences. They share one editorial discipline.

The lesson isn't "be like Mullvad on the website." Mullvad's competitive lane is narrower than the digital-asset projects, and their audience composition is different. The lesson is upstream of channel.

For digital-asset projects, the audience splits. Developers, traders, and early-adopter investors live on X, and the field serves them well. The audience that decides at scale, the institutional buyers, the integration partners, the regulators, the journalists writing background, the prospective hires evaluating culture, the search engines, the AI assistants that increasingly mediate "what should I look at" queries, doesn't decide on a viral thread. They decide after spending real time on the website and the blog. The field has built the first surface and underbuilt the second.

What the website is missing

When the methodology compares each project's narrative work across the five surfaces, four of the six dimensions of Narrative Strength consistently underperform on the website even though the project's other surfaces carry them strongly. The pattern isn't absence; the dimensions exist on the homepage. They're just not landing.

Narrative Completeness. Eight of fifteen projects underweight the website on Narrative Completeness, particularly on the friction beat from our 5-7-12 narrative blueprint. The homepage tells the reader what the project is, and less often what's at risk if the reader doesn't act. The customer is named, the path forward is implied, but the friction the customer is actually facing is left for the reader to fill in.

Message Quality. Six of fifteen projects underweight the website on Message Quality. The copy is informational. The headlines are descriptive. The proof is sparse. The voice doesn't quite have the texture that makes the reader feel the weight of what's on the line.

Conversion. Six of fifteen projects underweight the website on Conversion. The website doesn't answer "why now." The buyer-action moment is unmotivated, deferred, or simply absent.

Presence. Five of fifteen projects underweight the website on Presence, particularly on distinctiveness. The differentiation lives on the X feed and in the founder's posts. The homepage hides it behind generic category language.

These aren't exotic narrative moves. They're the basics of what a website is supposed to do for the reader who actually shows up to evaluate. The field has the material. The website isn't yet using it.

Compounding assets vs. perishable attention

There's an architectural distinction running underneath this finding that we use throughout the report.

A post does its work in 24 hours. A website page surfaces in search for 24 months. A blog post compounds for years if it earns its place. A canonical doc gets quoted by a reporter doing background two years from now. A case study gets referenced in a diligence memo. A podcast appearance lives in a directory for years. Compounding assets are the surfaces that keep working. Perishable attention is the surface that resets every morning.

For most projects, the website is the front door and the blog is the depth. They're the same compounding pair. Both are owned media. Both build psychological equity with the audience over time. The diligence team that finds a complete story on the homepage and a substantive blog is more likely to write the check or send the partnership email. The journalist doing background is more likely to take the project seriously. The search engine is more likely to surface the project to the next buyer in six months.

That's the audience the field has under-served. It's also the audience that compounds.

The new audience reads everything, every time

The audience that decides has grown. The institutional buyer is still there. The journalist is still there. The partnership lead is still there. There's a new reader now, and it reads differently than any reader before. It reads everything, every time, and it never sleeps.

That reader is AI.

The LLM mediating the next "what should I look at" query reads what you wrote, indexes it, and answers questions about your project to thousands of readers you will never meet.

For projects that built depth, this is amplification at scale. An AI explaining your project to a CFO at a Fortune 500 is paraphrasing your homepage. The clearer the story, the closer the paraphrase. For projects that didn't, this is silence at scale. The AI explains a competitor instead, or pulls a thinner picture from the X feed.

Most of the field hasn't built for that yet.

Finding 2

The privacy sector hasn't decided what privacy is for.

The deepest strategic split in the field sits inside one specific question: how should programmable privacy work, and what should it enable. Most individual projects have answered. The field has not.

When the methodology surfaces the most surprising sub-theme divergences inside that question, almost all of the surprises sit on a single fault line. Six projects on one side. Four on the other. The fault line cuts directly across the identity groups, which is what makes it the report's most strategic finding.

The compliance camp

Six projects sit in the compliance camp. Aleo, Canton, Miden, Midnight, Prividium, Zama. Each one positions privacy as the feature that enables institutional onboarding, regulator readiness, and compliant flows of value across a regulated network.

Aleo. The argument is "privacy as default, not a toggle." Aleo's case is that opt-in privacy fails because users default to convenience. The product behind the argument is USDCx, a privacy-first stablecoin built on Aleo mainnet with Circle's xReserve, plus USAD with Paxos. The compliance positioning has product behind it.

Miden. Pre-launch state with a specific institutional bet on stablecoins. The narrative is being pushed; the launch mechanic will determine whether it gets pulled by privacy press or by the broader scaling discourse. The compliance position is declared more than demonstrated, because the product isn't yet live.

Midnight. Compliance positioning with a Cardano sidechain anchoring, plus tokenized deposits via the Monument Bank partnership, plus an AI agent positioning that emerged in the recent window. The compliance bet is hybrid: institutional infrastructure plus AI-agent runtime. It's the most multi-vector bet inside the compliance camp.

Prividium. Sub-product of zkSync's enterprise stack. The "real economic utility" framing for the parent token positions Prividium as the ZK chain interop that makes Ethereum institutional finance work. The compliance position is consistent with zkSync's enterprise narrative; the question is whether Prividium builds the voice that owns "incorruptible institutional finance" directly or keeps letting zkSync and the analyst ecosystem carry the conversation.

Canton. The field's clearest pure-compliance position. Canton's sub-theme mapping is selective-disclosure-compliance at high confidence, with no composability emphasis on the other side at all. The Cantonomics tokenomics framework and the Circle xReserve integration with USDCx anchor the position in a tradfi-tokenization vector that's adjacent to the institutional-rails camp without splitting from it. Of every project in the compliance camp, Canton is the one whose mapping is anchored at the compliance pole rather than carrying a small composability lean as well.

Zama. Compliance camp on programmable privacy, universalizing on FHE. The most multi-thematic project in the field. Stablecoin infrastructure (the ERC-7984 confidential token standard, the first confidential stablecoin transfer on Ethereum). Principled launch (the sealed-bid Dutch auction conducted using the Zama protocol itself). Mainnet event. Technology demonstration. Developer programs. Five distinct narrative threads held coherently because the underlying argument is consistent: FHE is becoming the default, and Zama is showing what default looks like.

The composability camp

Four projects sit in the composability camp. Aztec, Oasis Network, Railgun, Zcash. Each one positions privacy as the feature that enables decentralized finance to actually be private, with composability across protocols.

Aztec. Soft composability lean. The sub-theme mapping places Aztec on the composability side without anchoring it as deeply as Oasis Network or Railgun. The recent commercial work (the Continuous Clearing Auction, the Ignition Chain mainnet, the privacy-by-default architecture via Noir) reads as developer-controllable privacy for programmable settlements rather than as institutional-rails infrastructure, which is consistent with the data position. Aztec is the closest project in the field to genuinely positioning across both camps; the data places it on the composability side by a thin margin.

Oasis Network. Composability camp on programmable privacy, with a specific application bet on AI-agent privacy via ROFL (Runtime Off-chain Logic). Two specific bets that both serve a non-institutional audience.

Railgun. "RAILGUN isn't a wallet. RAILGUN isn't a privacy coin. RAILGUN is infrastructure for resilient Ethereum privacy." The locked positioning copy is sharp. The narrative position is clear. The surfaces carrying it aren't yet at the depth the position deserves.

Zcash. Composability camp on programmable privacy, even though the cluster sections place Zcash in Rights as Refuge alongside WhatsApp by overall narrative shape. Zcash's specific bet is different from its cluster mate. The cluster name describes the project's identity. The camp name describes the project's direction. Zcash's direction is closer to Railgun and Oasis Network than to its cluster mate.

A methodology note on the placements

Two projects sit at unusual positions on the scale, and the methodology behind those placements is worth surfacing.

Aztec at the composability margin. Aztec's net composability emphasis is small, close to neutral. The data places Aztec on the composability side by a thin margin rather than anchoring it there as firmly as Oasis Network or Railgun. The signal is real, the magnitude is light. We treat Aztec as a soft composability lean.

Canton anchored at the compliance pole. Canton's mapping is pure selective-disclosure-compliance with no composability emphasis on the other side at all. Other compliance-camp projects (Aleo, Miden, Midnight, Prividium, Zama) carry small composability leans within their primary compliance position. Canton doesn't. That makes Canton the field's clearest pure-compliance anchor and explains why Canton sits at the far compliance end of the scale rather than in the middle.

Both projects have committed positions. One commits lightly, the other absolutely. The data places them where the data places them; we let the data lead.

Why this is strategic

Two camps. Two buyers. Two decks. The compliance camp is pitching central banks. The composability camp is pitching protocol designers. These audiences don't read the same press, don't attend the same conferences, don't write the same checks. A project that hasn't picked a side is asking both rooms for money with one slide deck. Both rooms notice.

The split connects to the conversion gap. A project that hasn't decided whether it is selling compliance or composability can't write a confident call to action. The "Restrained" projects on the discipline 2x2 are mostly compliance-camp. The "Conversion-fluent" projects (Brave, DuckDuckGo, Telegram) are consumer products with clear use cases. The Work to Do projects are upstream of both decisions. The strategic ambiguity at the camp level is feeding the asking ambiguity at the dimension level. If you don't know who you are selling to, your call to action doesn't know who to call.

Single-project signature emphases reveal positioning bets. Midnight on traditional-finance tokenization. Miden on compliant privacy stablecoins. Zama on FHE-as-blockchain-encryption. Oasis Network on privacy-rails-for-AI-agents. Railgun on infrastructure-for-resilient-Ethereum-privacy. These aren't interchangeable. The cluster sections give each project a sentence built on its specific bet rather than treating "Programmable Privacy" or "FHE Bet" as a uniform identity.

Finding 3

Discipline is binary.

Most variance in a field follows a normal distribution. Skill is normally distributed. Audience reach is normally distributed. Tenure, team size, talent depth, narrative intuition; all normally distributed across any reasonable cohort.

The privacy sector's Coherence and Conversion scores aren't.

When you look at how the cohort spreads across the six dimensions of Narrative Strength, four of them behave normally. Narrative Completeness, Signal Clarity, Message Quality, and Presence each vary continuously, the way you'd expect.

The remaining two look different. Coherence and Conversion are bimodal. A tight cluster of five projects sits near the top of each dimension, scoring 79 or higher. A separate cluster of ten sits in the middle, scoring 47 to 61. There's almost nothing between them.

That gap is the report's most important structural finding. The variance in the field on overall Narrative Strength is mostly variance on these two specific editorial choices a founder makes. Whether the story coheres in one sentence across surfaces. Whether the project asks the visitor for the next step.

Both are decisions, not capabilities. Skill, audience, reach are bell curves because each is shaped by a hundred factors. Coherence and Conversion are gaps because each is shaped by one. Has the founder decided who the project is? Has the founder decided what the visitor does next?

A project clears the gap or sits below it. Discipline is binary. You have decided. Or you have not.

The Coherence cluster

The five projects in the high-Coherence cluster are Aztec (81), Zama (81), Mullvad (79), Canton (82), and Aleo (82). The cluster isn't aligned with quadrant placement; it's aligned with editorial discipline. Two of these projects sit in Compounding (Canton and Zama). Two sit in Clear but Conventional (Aztec and Aleo). One sits in Compounding via the quiet-growth pattern (Mullvad). What they share isn't where they land on the chart. What they share is having decided who they are and held the decision across surfaces.

The Conversion gap

The Conversion cluster has a different shape. Brave (74), DuckDuckGo (75), and Telegram (70) anchor the high end as consumer products with clear use cases. The buyer-action moment is built into the product itself. Sign up. Download. Use it.

For the digital-asset projects, Conversion is the dimension where most fail. Of the eleven token-bearing projects we scored at the corpus level, eight score below 50 on Conversion. The project asks, but the ask doesn't motivate. Or the project doesn't ask at all.

The two dimensions aren't independent. The five high-Coherence projects all score above 50 on Conversion. The eight low-Conversion projects all score below 60 on Coherence. That's the report's deepest signal: when discipline is held, it's held across the dimensions that matter most for the buyer who arrives at the homepage cold.

What the three findings mean for each audience

For founders. Your cluster is your peer set. Your camp is your buyer. Your Coherence and Conversion scores tell you whether you've made the two binary decisions. If either sits in the lower cluster, the decision is the work.

For funds. Coherence and Conversion are the cleanest single-page diligence read in the report. A portfolio company in the lower cluster on either is a portfolio company that hasn't decided who it is or what it wants from the buyer. That's not a marketing problem. That's a strategy problem disguised as one.

For analysts and marketing leaders. A project that hasn't picked compliance or composability can't write a confident campaign. The campaign is downstream of the camp. Coherence tells you which projects have something quotable. The rest are still drafting.

Stage of life

Stage of life

Two phases sit inside the privacy sector at the same time. Some projects have broken through. Some are fighting to break through. The boundary between them is the most useful way to read the share-of-voice patterns that this section surfaces.

The three numbers

Three numbers tell the breakthrough story.

Total SOV. Your overall presence in the field's conversation. The full denominator: authored content plus third-party coverage as a percentage of the cohort total. The single number that says how loud you are.

Authored SOV. How much you are talking about yourself. The percentage of the field's authored content that comes from your project.

Third-party SOV. How much the world is talking about you. The percentage of the field's third-party coverage that discusses your project.

The ratio between authored and third-party is the breakthrough signal. A project producing far more authored voice than third-party coverage is competing on narrative because it hasn't yet broken through. A project receiving far more third-party coverage than authored voice has broken through; the world now carries the conversation, whether the project narrates it or not. A project where the two move in proportion is the rare balanced profile.

The trailing twelve months gives us the reading. Aleo (13.6% Total SOV, 98% authored) and Aztec (13.2% Total SOV, 98% authored) anchor the authored producers at substantial scale. Zcash (12.3% Total SOV, 92% third-party) anchors the press magnets. Canton (6.3% Total SOV, 75% authored, 25% third-party) and Prividium (1.8% Total SOV, 53% authored, 47% third-party) are the two balanced profiles. Brave (2.7% Total SOV, 89% authored) is the field's quiet growth pattern at scale: substantial authored voice, near-floor press, audience reach that lives outside the press cycle.

Three patterns. Three positions. Three different kinds of work.

The breakthrough phase

A handful of projects sit at the top of the field's third-party SOV of voice while producing very little authored content of their own. Their share of the conversation is dominated by what the world says about them, not by what they say about themselves.

Zcash. Approximately 35% of the field's third-party press, podcast, and news coverage over the trailing twelve months. Authored voice barely 1.5%. The ratio is the largest in the field by an order of magnitude. The Electric Coin Company team resigned in January 2026 citing constructive discharge by the Zcash Foundation board. The authored-voice silence is post-team-departure. External attention is heavily price-driven (ZEC up substantially year-over-year through 2025-2026). Plus institutional infrastructure (Foundry mining pool capturing nearly a third of hashrate in its first month). Plus security activity. Plus narrative debate (Charles Hoskinson's "I solved privacy" framing competing with ZK; quantum-threat discussions). The world is talking about Zcash. Zcash isn't talking about Zcash.

The Zcash voice arc is also the year's biggest. Total SOV moved from 2.27% in late August 2025 to a peak of 23.41% at end of December, retracing to 12.54% by March 31. A 21-point swing across four months tracks against the price run and the team-organizational cycle. The conversation is no longer accelerating. It's on the way back down, still five times where it started.

Telegram. Approximately 29% third-party SOV. Authored SOV approximately 1.5%. The third-party coverage is structural rather than strategic, driven by the Pavel Durov detention saga aftermath plus product feature press patterns. After the methodology adjustment for TON ecosystem coverage (which we treat as a separate non-cohort entity), Telegram's news volume holds at a level no other consumer-product project in the field reaches. Telegram has broken through at the corporate level. The TON ecosystem is breaking through separately.

Monero. Approximately 16% third-party SOV. Authored SOV under 1%. The third-party coverage is driven by regulatory pressure and price action. Authored voice is technical-only by core team practice (FCMP++ development, FerrySwap, the 12th anniversary), with zero token-marketing tweets in the window. Monero's posture is deliberate: the team has stayed anonymous and technical-voice-only since 2014. The market hasn't punished the silence yet (Monero's market position holds at the top of the privacy market by capitalization). Whether that pattern holds is the watchable signal for next cycle.

These three projects (plus WhatsApp at the platform-pull edge) have broken through. Their narrative work happens on a different plane. The question for them isn't "how do we build narrative discipline." The question is "how do we keep narrative compounding when the world is already carrying half the conversation, and when the conversation it is carrying may not match the project's own positioning."

The fighting-to-break-through phase

The other twelve to fourteen projects in the field produce substantially more authored voice than the world produces about them. The ratio of authored to third-party content across the field's recent window is roughly 2 to 1. Most of the field is publishing more than the world is picking up.

This isn't failure. It's the phase before the breakthrough.

Canton. The field's clearest balanced voice profile. Authored voice (approximately 7% of the field's authored SOV) plus proportional third-party press pickup (approximately 5%). The Cantonomics tokenomics framework launch and the Circle xReserve integration are pulling proportional trade-press coverage. Canton is the field's clearest case of authored voice translating to trade-press attention at the level the work warrants.

Midnight. Mainnet launch in March 2026 generated coverage and authored push together. The most balanced launch profile in the field, with the recency weighting capturing the surge cleanly.

Zama. Authored work substantial. Third-party press pickup hasn't yet caught up to the level the work warrants. The auction-via-FHE pattern, the ERC-7984 standard, the FHE mainnet milestone are all editorially substantive. The trade-press cycle hasn't yet found Zama at the cadence the work deserves.

Aztec. The token sale and the Ignition Chain mainnet generated some external coverage. The operations narrative (validator counts, decentralization proof, TVL growth) hasn't yet pulled press at the partnership scale Canton is operating at.

Aleo. USDCx launched with Circle on mainnet. The institutional press hasn't yet covered the partnership at the level it warrants.

Brave. The field's quiet growth case at substantial scale. Tweet engagement runs at over a third of the field's total tweet engagement (the "blocking ads in Brave is still $0 btw" tweet pulled 148,000 engagements). Third-party press pickup runs at approximately 1%. Brave has built consumer audience without dominating press conversation. We treat that pattern in detail in the adoption signal section below.

Mullvad, Proton. Activist voices that don't get amplified by trade press. Their audiences are advocates, journalists writing about surveillance, and policy adjacents. They convert through their own channels, not through press pickup. Their authored voice is louder than their third-party coverage by design.

Miden. Pre-launch state. Authored voice substantial; third-party privacy-sector coverage is essentially zero. Miden gets covered in non-privacy contexts (Polygon ecosystem, scaling discourse) that fall outside this report's scope. The launch mechanic chosen will shape which trajectory Miden gets.

Railgun. Position is clear; surfaces carrying the position are thin. Voice modest; third-party coverage modest.

These projects are doing the work that earns the breakthrough. Their authored voice exceeds their third-party coverage because they are competing on narrative because they haven't yet broken through. For some of them (Canton, Midnight, Aztec, Zama), the trade-press is starting to catch up. For others, the work is still being built.

Adoption signal alongside narrative position

We ran a parallel pass on public adoption data across the 18 projects to test whether the breakthrough framing holds against external signal. The metrics are heterogeneous by project type. Several figures are approximate or industry-estimated. The signal is directional. The full per-project table sits in the appendix.

Three findings sharpen the breakthrough framing.

Broken-through projects have adoption preceding their share of voice. Zcash carries roughly $5.4B market cap plus about a third of the network's mining hashrate via Foundry's pool plus ongoing institutional accumulation. Telegram operates at roughly 950M monthly active users. Monero carries approximately 58% of privacy-coin market capitalization. WhatsApp operates at roughly 3B monthly active users. The adoption signal precedes the share of voice. These projects didn't become talked about for hype reasons. They became talked about because their adoption was already substantial.

Several fighting-to-break-through projects are doing adoption work in the same window as their narrative discipline. Aztec shipped the Ignition Chain mainnet in March 2026 and reached approximately $1.2B in TVL within weeks. Zama produced the first confidential stablecoin transfer on Ethereum and the ERC-7984 confidential token standard. Canton's selective-disclosure stablecoin rails went live with Circle's xReserve. Aleo launched USDCx with Circle and USAD with Paxos. Midnight launched the NIGHT mainnet with federated node operators and tier-1 exchange listings. Five different shapes of the same pattern. Their narrative discipline isn't separate from their adoption work. It's the surface of it.

Brave is the field's clearest case of quiet growth at substantial scale. The browser carries roughly 85M monthly active users, the field's highest tweet engagement share by a wide margin, and a third-party press coverage share at near-floor levels.

The trajectory across time

The breakthrough divide isn't static. Voice allocation moves over time. The field's recent window includes major commercial events (mainnet launches, token auctions, stablecoin partnerships, organizational transitions) that shifted who was producing voice and who was receiving attention.

Reading the trajectory tells you which projects are in active commercial-event mode (the spike), which projects are sustaining (the plateau), and which projects are in transition (the inflection). Midnight's spike is the mainnet launch. Zcash's third-party rise is the price action plus the team-organizational event. Canton's sustained line is the tokenomics-and-stablecoin work landing as continuous narrative weight rather than as event-driven spikes.

The inverse pattern shows up too. Aztec, Telegram, and Oasis Network all show declining SOV over the trailing twelve months, not because absolute output dropped but because Zcash, Monero, and Canton rose. SOV is relative. A project that holds steady while others spike is losing share without losing volume.

The token narrative

The token narrative

The most important thing the digital-asset side of the field communicates about, after the product itself, is the token. Every project in this cohort with a live token markets that token directly. The token isn't a story riding inside the larger positioning narrative. It's a chapter the project writes about explicitly, by ticker, frequently, alongside everything else.

The data shows it cleanly. Volume varies by project. Zama mentions ZAMA in nearly half of its authored content. Aleo and Aztec each cross 22%. Miden, pre-launch, mentions MIDEN in 31% of posts. Even Brave at 2% has BAT in real volume. Six projects in the cohort don't market a project-issued token at all (Mullvad, Proton, Signal, DuckDuckGo, WhatsApp, Telegram corporate), and that absence is its own posture.

The strategic question isn't whether the field talks about tokens. It's where the token belongs in the narrative.

Where the token belongs

When the field discusses certain themes, the project's own token is in the same post most of the time. When it discusses other themes, the token is structurally absent. The split is the most strategically useful pattern in this section.

The token belongs here. When projects argue that transparent blockchain is a liability, the token comes into the conversation in two-thirds of the posts. The argument and the token are inseparable: if exposed payroll and visible treasury flows are the problem, the protocol-layer privacy that fixes them is the token's product. When projects discuss the programmable privacy stack or the FHE bet, the token shows up in roughly four out of ten posts. The architectural choice and the token are the same bet. When projects engage the illicit use reputation debate or the privacy coin regulatory pressure themes, the token is in nearly half of the posts because the regulatory conversation IS about the token. When projects discuss institutional onchain privacy rails, the token shows up in a quarter of posts, because the rails are token rails.

The token doesn't belong here. When projects discuss the surveillance economy positioning (consumer alternative to Big Tech), the token is in fewer than one in five posts. When they invoke privacy as a fundamental right or push back on surveillance state legislation, the token shows up in roughly one in eight posts. When they argue messaging apps as distribution layer for digital assets, the project's own token shows up 1.6% of the time. Across the consumer, rights, and state-actor themes, the token is structurally absent from the narrative. These are conversations the field has without bringing the token in.

The strategic implication is clear. Token narrative compounds in architectural, financial, and transactional conversations. It does not compound in consumer-values or rights-framing conversations. A project pitching the institutional buyer can tie the token to the pitch. A project pitching the rights-aligned consumer is competing against itself if it tries to tie the token to the pitch.

The five jobs the token talk does

Across the projects that mention the token, the talk does five jobs. Each is a strategic shape. Each project leans on different ones.

Utility. What holding the token does for the holder. Payments, staking, governance, validator participation, ecosystem access. The most common job. Aleo connects ALEO to private payments via the Shield wallet and the Aleo Card. Brave connects BAT to ad-blocking rewards. Oasis connects ROSE to confidential compute access. Utility is the only job that grounds the token in something the holder actually does today.

Does the utility narrative ground in product the holder can use, or in roadmap?

Launch event. Mainnet, sale, TGE, generation event. Concentrated at the active-launcher projects: Zama (auction-via-FHE), Aztec (Continuous Clearing Auction plus Ignition Chain), Aleo (USDCx, USAD on mainnet). Done well, the launch IS the proof.

What does the launch sound like when it's the project's positioning argument made operational, instead of a financial event with a brand campaign next to it?

Product event. Integrations, deployments, real-world partnerships. Aleo and Aztec lead this register. Each product release doubles as a token statement because the product runs on the token.

Does each product release reinforce the token's utility narrative, or does product run in a track parallel to the token?

Distribution. Allocations, fellowships, grants, airdrops. Midnight stands out at 51% of its token-mention posts. Distribution is recognition of contribution. Token allocations to early developers, ecosystem partners, content creators, and technical reviewers build the audience that compounds.

Are distribution allocations going to people who showed up before the project had to pay them, or to liquidity sources that exit at first listing?

Validator economics. Validator rewards, delegation, consensus economics. Aztec, Miden, and Canton lean here. The most institutional-finance-legible of the token jobs.

Does the validator economics narrative speak to the buyer the project is actually pitching, or to a different buyer the project is signaling toward?

Volume isn't the question. Integration is.

The strategic frame this section is really measuring is whether direct token marketing connects to the project's positioning narrative or runs parallel to it.

Four projects integrate at 65% or higher: Aleo (70.6%), Midnight (67.3%), Zama (67.1%), Miden (66.9%). When they mention the token, they almost always connect it to the positioning argument in the same post. The token rides inside the narrative, not parallel to it.

Aztec sits in the middle at 49.1% integration despite the highest direct-token volume in the field after Zama. The most volume in the field doesn't translate to the most integrated marketing. The diagnostic: AZTEC the token is being talked about at scale, but only half the time tied back to the architectural and launch positioning the project is otherwise carrying.

Three projects sit at the bottom: Oasis Network (33%), Monero (40%), Railgun (20%). Each for a different reason. Oasis produces the largest authored corpus in the cohort but rarely names ROSE (2.3%), and when it does, the token isn't tied to positioning. Monero's posture explains its low everything. Railgun has the sharpest positioning copy in the field but doesn't put RAIL into the conversation in any meaningful volume.

Methodology

Methodology and appendix

The methodology in this report is descriptive of public surfaces these projects already publish on. We're reading the work the field has done in public. We don't measure product quality, team quality, technology quality, market traction, revenue, or token price. Those signals exist elsewhere. What we're measuring is narrative discipline.

The report rests on a sequence of analytical layers, each feeding the next.

1. Content ingest. We ingest privacy-field content across five mediums (website, blog, news, podcast, X) for a 365-day window. Project-sourced classification distinguishes what each project is saying about itself from what the world is saying about the project. The ingest combines firehose data from The Tie with custom F/AS scrapers and project-specific blog rules.
2. Cohort taxonomy. We extract the field's narrative-positioning taxonomy as 12 root themes and 22 sub-themes directly from the corpus, validated against per-project coverage and synthesis-quality gates before promotion.
3. Per-project mapping. Each project gets mapped to the taxonomy with confidence scores and evidence references. Two post-processing rules tighten the mapping: rolling sub-theme mappings up to their parent root themes when justified, and capping confidence on mappings that conflict with cohort-level signal.
4. Cross-channel consistency. We measure whether each project's narrative is consistent across mediums or scatters across surfaces. Channel-shift flags surface when a project's recent voice migrates to a different dominant medium than its cumulative voice.
5. Alignment and clustering. We score each project's alignment with the field's narrative center, then cluster projects into identity groups based on narrative shape.
6. Narrative Strength scoring. The headline metric. Each project gets evaluated cell by cell across 145 cells (29 narrative classifications across 5 mediums). The classification grid feeds a base score; a bounded quality adjustment refines the base score based on cross-medium pattern. Three projects could not be scored at the corpus level and substitute a website-only score.
7. Token narrative. For projects with tokens, a parallel taxonomy surfaces how each project integrates token narrative into other registers. The token narrative analysis is sector-bespoke; the registers are the costumes the privacy sector's token narrative wears.
8. Share of voice. Per-project Authored SOV, Third-party SOV, and Total SOV, recency-weighted across the year, with engagement-weighted secondary view for posts on X.

Recency framework. Content from the last 60 days carries full weight. Content from 61-180 days carries half weight. Content from 181-365 days carries quarter weight. Older content is filtered. The framework is consistent across all layers.

The full methodology is documented in a separate IP-controlled document. Analysts who want methodology trust before citing the report can request access under partnership conversation with appropriate mutual NDA.

What the report can't tell you

The methodology measures public surfaces. It doesn't measure product quality, team quality, technology quality, market traction, revenue, token price, user retention, or any signal that requires private data.

The field is 18 surviving privacy projects. Projects that died because nobody understood them aren't in the dataset. Findings are descriptive of survivors.

Cluster averages drawn from one or two projects (Programmable Messenger, Hard Privacy, the FHE Bet, Rights as Refuge) are editorial readings, not statistical claims. Programmable Privacy at seven projects and Anti-Surveillance Internet at five projects are the only clusters where averages carry weight.

Sentiment data is mixed across mediums. Cross-medium sentiment comparisons aren't used in this report.

We aren't claiming Narrative Strength predicts market price. The forward-looking question of whether discipline translates to market position is watchable for next cycle.

F/AS provides positioning and messaging strategy. The methodology uses public surfaces these projects already publish on. The diagnosis is descriptive.

Substituted scoring

Three projects substitute a website-only score for the corpus-level Narrative Strength. Each substitution has a different reason, and each reason is editorial signal in itself.

Signal. Insufficient corpus by editorial design. Authored voice is sparse by design. Quiet from a project this clear is discipline, not deficit.

Prividium. Insufficient corpus due to brand subordination. Coverage is mostly through zkSync ecosystem analyst chatter. The market is louder than the project is.

Proton. The safety classifier on the methodology refused on policy-sensitive content. The Proton blog carries investigative-journalism-grade activism. The website-only methodology returned a clean score; the corpus-level methodology returned a refusal. The duality is the discipline.

Glossary of key terms

Compliance camp. Projects whose programmable-privacy emphasis sits on selective-disclosure-compliance, serving institutional buyers, regulators, and tradfi onboarding.

Composability camp. Projects whose programmable-privacy emphasis sits on private decentralized-finance composability, serving DeFi protocol designers and traders.

Compounding asset vs. perishable attention. Architectural distinction between surfaces that hold value across time (website, canonical doc, case study, podcast) vs. surfaces that reset each morning (the X feed).

Differentiation. The horizontal axis on the signature 2x2. Distance from the field's narrative center of gravity. High differentiation means the project's narrative position is distinctive relative to peers. Not a quality measure.

Token narrative integration. Whether direct token marketing connects to the project's positioning narrative within the same authored content. Measured as the percentage of token-mention posts that also reference positioning themes (privacy, institutional, composability, technical primitives). High integration means the token rides inside the positioning narrative; low integration means the token campaign runs parallel to it.

Identity group. One of the six clusters in the field. Describes who a project is to its peer set.

Narrative Strength (NS). F/AS proprietary score, 0-100, weighted across six dimensions. The vertical axis on the signature 2x2.

No-token-by-design. A project that doesn't market a token of its own. Six projects in the field: DuckDuckGo, Mullvad, Proton, Signal, Telegram (corporation), WhatsApp.

Quiet growth. A pattern where a project builds substantial audience without dominating press conversation. Brave at approximately 85M MAU with near-floor third-party press coverage is the field's clearest example at scale.

Substituted score. A project scored with a website-only methodology instead of the corpus-level methodology. Three substitutions: Signal, Prividium, Proton.

Voice allocation. Per-project distribution of authored voice across registers, recency-weighted.

Credits and contact

F/AS. fas.partners. Strategic advisory for technical founders, with concentration in digital assets, AI, and infrastructure. Author and contact: Devon Ferreira, devon@fas.partners. Questions, corrections, and factual disputes welcomed.

The team. Methodology stack designed and operated by F/AS. Atticus on prose. Soren on methodology and rigor. Clara on assembly and editorial through-line. Maren on information hierarchy. Cornelius on orchestration. Kenji on technical architecture. Monty on legal review. Glen on strategic framing.

The Tie. Special thanks to The Tie for the firehose data feed (tweet, news, podcast) that anchors the corpus, and for the partnership that made this report's measurement architecture possible at the fidelity it needed.

Data sources. The Tie (firehose for tweet, news, podcast). F/AS custom scrapers (website, blog, podcast extension). Public market data via standard digital-asset tracking sources accessed 2026-05-01.

Report assembled. May 2026. Saturday 2026-05-02 publish at fas.partners/intelligence/privacy.

Methodology rigor. The full methodology is documented in a separate IP-controlled document. Available under partnership conversation with appropriate mutual NDA.

The privacy sector spent eighteen months building. This report measures the work that decides whether the building compounds. We read public surfaces. We didn't grade product, team, or technology. We graded the story.

Your story is the only thing the audience has read of you. Make it carry the weight.

Schedule a working session and walk-through with Devon.