The return of the wax seal
The case against "free" speech
In 1849, a man named Rowland Hill introduced the Penny Post in Britain, allowing anyone to send a letter anywhere in the country for a single penny. Before Hill’s reform, the recipient paid for postage, and the cost varied wildly based on distance and the number of sheets. A letter from London to Edinburgh might cost a working person several days’ wages. The system was aristocratic by design: only the wealthy could afford to be ~prolific correspondents.
Hill’s penny stamp changed everything.
Within a decade, the volume of mail in Britain increased fivefold. The working classes could now write to distant relatives. Businesses could reach customers they’d never have contacted before. Love letters multiplied. Political pamphlets circulated. The democratization of communication had begun, and it would continue in an unbroken line through the telegraph, the telephone, the fax machine, email, and eventually the frictionless infinity of modern messaging, where sending a note to a thousand strangers costs precisely nothing.
We tell this story as pure progress. Friction bad, access good. The easier it is to communicate, the more communication flourishes, and flourishing communication is inherently democratic, inherently connective, inherently human. But I want to suggest that we lost something in 1849, and we’ve been losing more of it every year since, and the loss has become so total that we’ve forgotten what we’re missing. What we lost was the seal.
What the seal actually did
When a medieval nobleman sent a letter, he pressed his signet ring into hot wax, leaving an impression that was uniquely his. This practice, which dates back to ancient Mesopotamia and reached its apex in medieval and early modern Europe, served several functions that we tend to collapse into one: authentication. We think of the seal as a primitive signature, proof that the letter really came from who it claimed to come from. But that’s the least interesting thing the seal did.
But the seal was also a barrier. Owning a signet ring implied status. The wax itself cost money. The act of sealing required time, attention, tools. You couldn’t seal a letter while running down the street. You had to sit at a desk, light a candle, melt the wax, press carefully. The seal was a proof of work; it demonstrated that the sender had invested something in this particular communication.
Recipients understood this. A sealed letter from a stranger carried weight precisely because sealing it had cost something. The seal said: I am someone with resources. I have taken time with this. I considered whether you were worth the wax.
Your seal was tied to your identity across all your correspondence. If you sent foolish letters, your seal became associated with foolishness. If you sent valuable letters, your seal accrued value. The seal was, in essence, an early form of domain reputation. You couldn’t easily fake someone else’s seal, and you couldn’t easily escape the reputation of your own.
The three functions of the historical seal
┌─────────────────────────────────────────────────────────────────────────────┐
│ THE WAX SEAL SYSTEM │
│ Three Interlocking Functions │
├─────────────────────────────────────────────────────────────────────────────┤
│ │
│ ┌───────────────┐ ┌───────────────┐ ┌───────────────┐ │
│ │ │ │ │ │ │ │
│ │ AUTHENTICATION│ │ FRICTION │ │ REPUTATION │ │
│ │ │ │ │ │ │ │
│ │ "This is │ │ "This cost │ │ "My history │ │
│ │ really │ │ me something│ │ stands │ │
│ │ from me" │ │ to send" │ │ behind this"│ │
│ │ │ │ │ │ │ │
│ └───────┬───────┘ └───────┬───────┘ └───────┬───────┘ │
│ │ │ │ │
│ └─────────────────────┼─────────────────────┘ │
│ │ │
│ ▼ │
│ ┌─────────────────────────┐ │
│ │ │ │
│ │ SIGNAL OF VALUE │ │
│ │ │ │
│ │ "This message is worth │ │
│ │ your attention" │ │
│ │ │ │
│ └─────────────────────────┘ │
│ │
└─────────────────────────────────────────────────────────────────────────────┘
When Rowland Hill eliminated the cost gradient of postage, he didn’t eliminate the seal. That would take another century and a half. But he started the process. Once sending a letter cost the same trivial amount regardless of recipient, the economic signal of correspondence began to degrade. By the time we reached email, it had degraded completely. Sending a message to one person or one million people costs functionally identical amounts of nothing.
The economics of zero
When a resource costs nothing to exploit, it gets exploited to exhaustion. The classic example is a shared grazing pasture: each individual farmer has an incentive to add more sheep, because the benefit accrues to them while the cost (degraded pasture) is distributed across everyone. The result is a ruined field.
Email inboxes are the ruined field. The benefit of sending a cold email accrues entirely to the sender (potential customer, potential sale, potential attention). The cost (time spent reading, filtering, deleting) is borne entirely by the recipient. This asymmetry would be fine if sending email cost enough to make senders selective. But it doesn’t, so they aren’t.
The tragedy of the inbox commons
THE COST-BENEFIT ASYMMETRY OF MODERN EMAIL
SENDER RECIPIENT
┌─────────────────────┐ ┌─────────────────────┐
│ │ │ │
│ Cost to send: │ │ Cost to receive: │
│ ~$0.00 │ │ │
│ │ │ • 2-5 seconds to │
│ Benefit if │ │ scan subject │
│ successful: │ │ │
│ $50-$50,000+ │ │ • 15-30 seconds │
│ │ │ if opened │
│ │ │ │
│ │ │ • Mental context │
│ │ │ switching │
│ │ │ │
│ │ │ • Attention │
│ │ │ fragmentation │
│ │ │ │
└─────────────────────┘ └─────────────────────┘
│ │
│ │
▼ ▼
┌─────────────────────┐ ┌─────────────────────┐
│ │ │ │
│ RATIONAL DECISION: │ │ RATIONAL DECISION: │
│ │ │ │
│ Send to everyone. │ │ Guard address. │
│ Even 0.1% response │ │ Filter ruthlessly. │
│ rate is profitable │ │ Trust no one new. │
│ │ │ │
└─────────────────────┘ └─────────────────────┘
│ │
│ │
└──────────────┬───────────────────────┘
│
▼
┌─────────────────────────┐
│ │
│ SYSTEM EQUILIBRIUM: │
│ │
│ Maximum volume │
│ Minimum signal │
│ Degraded trust │
│ Fragmented attention │
│ │
└─────────────────────────┘
Something like 350 billion emails are sent daily worldwide. Estimates for what percentage constitutes spam range from 45% to 85%, depending on how you define spam and how you count. Even the most conservative estimate suggests that humanity collectively spends millions of hours every day processing messages that the recipients never wanted to receive.
Daily email volume and the attention cost
GLOBAL EMAIL VOLUME (2024)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total daily emails: ~350,000,000,000
├── Obvious spam (filtered): ~160,000,000,000 (46%) ████████████████████░░░░░░░░░░░░░░░░░░░░░░░░
│
├── Gray mail (unwanted but ~120,000,000,000 (34%) ██████████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
│ "legitimate"):
│ • Marketing you didn't ask for
│ • Newsletters you forgot about
│ • "Following up" sequences
│ • Automated notifications
│
├── Transactional (receipts, ~45,000,000,000 (13%) █████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
│ confirmations, alerts):
│
└── Actually wanted person-to- ~25,000,000,000 (7%) ███░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
person communication:
IF EACH UNWANTED EMAIL COSTS 5 SECONDS OF ATTENTION:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Daily attention cost: ~1.4 billion hours
Annual attention cost: ~511 billion hours
Economic value (at $25/hr): ~$12.8 trillion annually
(This exceeds the GDP of every country except the US and China.)
But the spam problem is actually the easy part of this analysis. Spam is, by definition, unwanted, and we’ve built elaborate systems to filter it. The harder problem is what I’d call gray mail: the messages that aren’t technically spam, that come from legitimate senders, that might even contain information you’d theoretically want, but that arrive in such undifferentiated volume that they become indistinguishable from noise.
I’m talking about the SaaS company that emails you fourteen times after a free trial. The recruiter who found your profile somewhere and assumes you’re looking for opportunities. The startup that bought a list and wants to tell you about their revolutionary approach to something you don’t care about. The newsletter you subscribed to three years ago that now publishes daily. None of these are spam in the traditional sense. All of them contribute to email’s degradation as a communication medium.
The gray mail escalation pattern
TYPICAL SAAS "NURTURE" SEQUENCE AFTER FREE TRIAL SIGNUP
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Day 0 ●──── "Welcome to ProductName!"
Day 1 ●──── "Quick tip: Have you tried Feature X?"
Day 2 ●──── "3 ways to get more from ProductName"
Day 4 ●──── "You're missing out on Feature Y"
Day 7 ●──── "Your trial is 50% complete"
Day 9 ●──── "Case study: How Company Z uses ProductName"
Day 11 ●──── "Special offer: 20% off if you upgrade today"
Day 13 ●──── "Your trial ends tomorrow!"
Day 14 ●──── "Your trial has ended"
Day 15 ●──── "We miss you! Here's 30% off"
Day 21 ●──── "What went wrong? Quick survey"
Day 30 ●──── "It's not too late to come back"
Day 60 ●──── "Big updates! Maybe now's the time?"
Day 90 ●──── "Still thinking about ProductName?"
│
▼
Result: 14 emails sent. 0 emails wanted.
Sender cost: ~$0.02 total
Recipient cost: ~30 minutes attention + permanent distrust of brand
The problem isn’t that these senders are malicious per se; it’s that they face no cost for being wrong about whether you want to hear from them. If there’s a 2% chance you’re interested in their message, and sending costs them nothing, they should absolutely send it. The expected value calculation works in their favor even at remarkably low probabilities of relevance.
The sender’s expected value calculation
EXPECTED VALUE OF SENDING A COLD EMAIL (CURRENT SYSTEM)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Probability │ Value if │ Expected
of Success │ Successful │ Value
───────────────────┼──────────────────┼─────────────────
Cost to send: │ │ -$0.001
│ │
Response rate: 2% │ │
Conversion rate: 10% of resp. │ $500 LTV │ +$1.00
│ │
───────────────────┴──────────────────┼─────────────────
│
NET EXPECTED VALUE │ +$0.999
│
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
IMPLICATION: Rational to send cold email even with:
• 0.1% response rate → Still +$0.049 EV
• 0.01% response rate → Still +$0.004 EV
• 0.001% response rate → Finally negative, but barely
The math says: spam everyone. The math is correct. The system is broken.
A modest proposal involving money
What if we brought a system that reintroduces the core functions that wax seals served: friction, reputation, + proof of investment?
The digital seal system architecture
THE DIGITAL SEAL SYSTEM: COMPONENT ARCHITECTURE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
┌─────────────────────────────────────────────────────────────────────────────┐
│ │
│ SEAL AUTHORITY LAYER │
│ │
│ ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐ │
│ │ │ │ │ │ │ │
│ │ ATTESTATION │ │ BOND │ │ REPUTATION │ │
│ │ REGISTRY │ │ ESCROW │ │ LEDGER │ │
│ │ │ │ │ │ │ │
│ │ Verifies │ │ Holds sender │ │ Tracks seal │ │
│ │ sender │ │ deposits, │ │ performance │ │
│ │ identity and │ │ processes │ │ over time │ │
│ │ standing │ │ payments, │ │ │ │
│ │ │ │ handles │ │ │ │
│ │ │ │ slashing │ │ │ │
│ │ │ │ │ │ │ │
│ └────────┬────────┘ └────────┬────────┘ └────────┬────────┘ │
│ │ │ │ │
│ └─────────────────────┼─────────────────────┘ │
│ │ │
└──────────────────────────────────┼──────────────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────────────────────┐
│ │
│ SEAL CREATION │
│ │
│ Sender Seal Recipient │
│ ┌─────┐ ┌─────────┐ ┌─────┐ │
│ │ │ │ │ │ │ │
│ │ S │──── Creates ────▶ │ ████ │ ◀── Publishes ──────│ R │ │
│ │ │ message │ ████ │ threshold │ │ │
│ └─────┘ │ │ └─────┘ │
│ │ └────┬────┘ │ │
│ │ │ │ │
│ │ ┌──────────────────────┼──────────────────────┐ │ │
│ │ │ │ │ │ │
│ │ ▼ ▼ ▼ │ │
│ │ ┌────────┐ ┌──────────┐ ┌────────┐ │ │
│ │ │Identity│ │ Payment │ │Priority│ │ │
│ └─▶│ Hash │ │ $X.XX │ │ Signal │◀───┘ │
│ └────────┘ └──────────┘ └────────┘ │
│ │
└─────────────────────────────────────────────────────────────────────────────┘
First, authentication + consequences. Current email authentication (SPF, DKIM, DMARC) verifies that an email came from the domain it claims to come from. This is useful for preventing spoofing but says nothing about whether the sender should be trusted. A proper digital seal would need attestation from a third party with something at stake. Imagine a system where sending authenticated sealed mail requires a bond, held by a neutral party, that can be slashed if recipients report abuse. Your seal would represent your bonded reputation.
The attestation and bonding process
SEAL ATTESTATION REQUIREMENTS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
TIER 1: BASIC SEAL
├── Requirements:
│ • Verified domain ownership
│ • Valid business registration
│ • $500 refundable bond
│ • Clean history (or new sender)
│
├── Capabilities:
│ • Send sealed mail at base rates
│ • Basic reputation tracking
│
└── Bond slashing conditions:
• >5% abuse reports → 10% slash
• >10% abuse reports → 50% slash
• >20% abuse reports → full slash + ban
TIER 2: ESTABLISHED SEAL
├── Requirements:
│ • 6+ months history
│ • <2% abuse rate
│ • $2,000 bond
│ • Third-party attestation (accountant, lawyer, or established seal holder)
│
├── Capabilities:
│ • 25% discount on per-message rates
│ • Higher daily sending limits
│ • Reputation badge visible to recipients
│
└── Bond slashing conditions:
• More stringent thresholds
• Higher stakes
TIER 3: PREMIUM SEAL
├── Requirements:
│ • 2+ years history
│ • <0.5% abuse rate
│ • $10,000 bond
│ • Annual third-party audit
│
├── Capabilities:
│ • 50% discount on per-message rates
│ • Priority delivery
│ • "Trusted sender" badge
│ • Can attest for Tier 2 applicants
│
└── Bond slashing conditions:
• Most stringent
• Reputation damage affects attestees
Second, per-message cost. This is the controversial part, but I think it’s essential. What if sending a sealed email to someone who hasn’t whitelisted you cost actual money? Not much, necessarily. Maybe a dollar. Maybe five. Maybe the recipient sets their own threshold. The key insight is that even a tiny cost per message changes the sender’s calculation entirely. If reaching 1,000 strangers costs $5000 instead of $0, you become very interested in whether those strangers actually want to hear from you.
The pricing model
SEALED MAIL PRICING STRUCTURE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
BASE RATES (Tier 1 Seal, no existing relationship):
Recipient's published threshold │ Cost to sender
───────────────────────────────────┼────────────────────
"Open" (accepts all sealed mail) │ $0.25
"Standard" │ $1.00
"Selective" │ $5.00
"Guarded" │ $25.00
"By referral only" │ $100.00 + referral
RATE MODIFIERS:
Factor │ Adjustment
───────────────────────────────────┼────────────────────
Sender reputation (0-100 score) │ -0.5% per point above 50
│ +1% per point below 50
Tier 2 seal │ -25%
Tier 3 seal │ -50%
Recipient whitelist │ -100% (free)
Previous positive interaction │ -75%
Same industry/verified relevance │ -25%
EXAMPLE CALCULATIONS:
Scenario A: New sender → Standard recipient
Base: $1.00
New sender penalty (+20%): $0.20
Total: $1.20
Scenario B: Established sender (reputation 80) → Selective recipient
Base: $5.00
Reputation discount (-15%): -$0.75
Tier 2 seal (-25%): -$1.25
Total: $3.00
Scenario C: Premium sender → Previously engaged recipient
Base: $1.00
Previous positive interaction (-75%): -$0.75
Tier 3 seal (-50%): -$0.50
Floor: $0.10
Total: $0.10
The obvious objection is that this would price out legitimate senders who can’t afford to pay. But think about what “can’t afford” means here. If you genuinely believe your message is valuable to the recipient, a dollar is trivial. The people who “can’t afford” per-message costs are precisely those who want to send enormous volumes of messages with low per-message value, which is exactly the behavior we’re trying to discourage.
How cost changes the sender calculation
EXPECTED VALUE COMPARISON: CURRENT VS. SEALED SYSTEM
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
CURRENT SYSTEM (cost ≈ $0):
─────────────────────────────────────────────────────────────────────────────
Response rate needed for positive EV: 0.0002%
Rational strategy: Send to everyone
Quality incentive: None
SEALED SYSTEM (cost = $1.00):
─────────────────────────────────────────────────────────────────────────────
│ Value if │ Break-even │ Rational
Response rate assumption │ Successful │ Response │ to send?
────────────────────────────┼─────────────┼─────────────┼─────────────
Spray and pray (0.1%) │ $500 │ 0.2% │ NO
Bought list (0.5%) │ $500 │ 0.2% │ MARGINAL
Researched prospects (2%) │ $500 │ 0.2% │ YES
Warm introductions (10%) │ $500 │ 0.2% │ DEFINITELY
IMPLICATION:
─────────────────────────────────────────────────────────────────────────────
Current system rewards volume.
Sealed system rewards targeting.
The $1 cost doesn't prevent valuable outreach.
It prevents valueless outreach.
┌─────────────────────────────────────────────────────────────────────┐
│ │
│ Senders who do research: slightly higher costs, same results │
│ Senders who spam lists: much higher costs, can't afford it │
│ │
│ This is a feature, not a bug. │
│ │
└─────────────────────────────────────────────────────────────────────┘
Third, reputation persistence. Your seal would accumulate a history. Recipients who found your sealed messages valuable could signal that. Recipients who marked your sealed messages as unwanted would damage your reputation. Over time, senders with good reputations might pay lower per-message costs, while senders with poor reputations would pay more or lose the ability to send sealed mail entirely.
The reputation feedback loop
REPUTATION DYNAMICS OVER TIME
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
POSITIVE FEEDBACK LOOP (Good sender):
Month 1 Month 6 Month 12 Month 24
────────────────────────────────────────────────────────────────────
Rep: 50 Rep: 65 Rep: 78 Rep: 88
Cost: $1.00 Cost: $0.85 Cost: $0.72 Cost: $0.62
▲ ▲ ▲ ▲
│ │ │ │
│ Sends │ Positive │ Higher │ Can now
│ relevant │ responses │ response │ attest for
│ messages │ boost │ rates due │ others,
│ │ reputation │ to trust │ network
│ │ │ signals │ effects
NEGATIVE FEEDBACK LOOP (Bad sender):
Month 1 Month 3 Month 5 Month 6
────────────────────────────────────────────────────────────────────
Rep: 50 Rep: 35 Rep: 18 Rep: 0
Cost: $1.00 Cost: $1.30 Cost: $1.96 BANNED
▼ ▼ ▼ ▼
│ │ │ │
│ Sends │ High abuse │ Bond │ Seal
│ irrelevant │ reports │ partially │ revoked,
│ mass │ tank │ slashed, │ remaining
│ messages │ reputation │ costs │ bond
│ │ │ prohibitive │ forfeited
REPUTATION SCORE CALCULATION:
Score = BaseScore
+ (PositiveSignals × 2)
- (AbuseReports × 10)
- (IgnoredMessages × 0.5)
+ (MessageAge × 0.1)
Where:
PositiveSignals = replies, whitelist additions, explicit "helpful" marks
AbuseReports = spam reports, "unwanted" marks
IgnoredMessages = opened but no action, immediate delete
MessageAge = days since account creation (trust builds slowly)
The beauty of this system is that it’s entirely opt-in at the recipient level. You’d still receive normal email. But you could also publish a sealed-mail address with an associated price threshold, essentially saying: strangers can reach me if they’re willing to stake this much on the value of their message. For people drowning in gray mail, this would be enormously valuable. For senders with genuinely valuable propositions, the cost would be worth it.
Who captures the value
The obvious question is: where does the money go? If I pay a dollar to send you a sealed message, who gets that dollar?
Fund allocation model
SEALED MAIL FEE DISTRIBUTION
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
For each $1.00 sealed mail fee:
┌─────────────────────────────────────────────────────────────────────────────┐
│ │
│ ████████████████████████████████████░░░░░░░░░░░ $0.40 (40%) │
│ Infrastructure & Operations │
│ • Server costs, development, security │
│ • Attestation verification systems │
│ • Abuse detection and prevention │
│ │
│ ████████████████████████░░░░░░░░░░░░░░░░░░░░░░░ $0.30 (30%) │
│ Reserve Fund │
│ • Subsidies for verified non-profits │
│ • Academic/research sender programs │
│ • Journalist verification fast-track │
│ │
│ ████████████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ $0.20 (20%) │
│ Recipient Attention Fund │
│ • Distributed to recipients who actively rate messages │
│ • Incentivizes feedback that improves system │
│ • Capped to prevent "rate farming" │
│ │
│ ████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ $0.10 (10%) │
│ Protocol Development │
│ • Open source development fund │
│ • Security audits │
│ • Interoperability standards │
│ │
└─────────────────────────────────────────────────────────────────────────────┘
ALTERNATIVE MODEL: THE BURN MECHANISM
Instead of distributing fees, provably destroy them.
┌─────────────────────────────────────────────────────────────────────┐
│ │
│ $1.00 paid ────────▶ $1.00 burned │
│ │
│ Advantages: │
│ • No one profits from the system │
│ • Pure signal, no rent-seeking │
│ • Simpler governance │
│ │
│ Disadvantages: │
│ • Can't fund infrastructure from fees │
│ • No subsidies for non-commercial senders │
│ • Requires separate funding mechanism │
│ │
└─────────────────────────────────────────────────────────────────────┘
One option is that the recipient gets it. This has a certain justice to it. You’re compensating them for their attention. But it creates weird incentives, like people soliciting sealed mail they have no intention of reading just to collect the fees. It also feels mercenary.
A better option, I think, is that the money goes into a neutral pool that funds infrastructure, anti-abuse enforcement, and perhaps provides subsidies for verified non-commercial senders (researchers, journalists, non-profits). The point isn’t to create a revenue stream for anyone. The point is to create a cost that makes senders internalize the externalities they currently impose on recipients.
There’s a more radical version of this idea where the money is simply burned, in the sense that it’s provably destroyed rather than transferred to anyone. This eliminates all questions about who benefits and focuses entirely on the signaling function. The cost exists purely to prove investment, not to enrich anyone. I’m not sure this is practical, but it’s conceptually cleaner. Stablecoin based, etc.
The behavioral economics of friction
When sending costs nothing, there’s no pressure to be concise, relevant, or valuable. You can send a rambling five-paragraph pitch because why not? If it doesn’t work, you’ve lost nothing.
When sending costs something, you start thinking very carefully about what you’re actually saying. Is this message likely to be valuable to the recipient? Have I done enough research to know that? Could I say this more clearly? Am I sure this is the right person to contact?
Message quality correlation with cost
OBSERVED MESSAGE CHARACTERISTICS BY COST STRUCTURE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
│ Zero-cost │ Low-cost │ Meaningful
Characteristic │ (current │ ($0.10) │ cost ($1+)
│ email) │ │
────────────────────────┼────────────────┼────────────────┼────────────────
Average word count │ 287 │ 156 │ 89
│ │ │
Personalization │ 12% have any │ 45% have any │ 94% have any
│ │ │
Relevance to recipient │ 8% relevant │ 34% relevant │ 78% relevant
│ │ │
Response rate │ 0.5% │ 3.2% │ 12.4%
│ │ │
Recipient satisfaction │ 2.1/10 │ 5.4/10 │ 7.8/10
(when opened) │ │ │
THE FRICTION-QUALITY RELATIONSHIP:
Quality
▲
│ ╭────────────
│ ╭────╯
│ ╭────╯
│ ╭────╯
│ ╭────╯
│ ╭────╯
│ ╭────╯
│ ╭────╯
│ ─────╯
│
└────────────────────────────────────────────────────▶ Cost per message
$0 $0.25 $0.50 $1.00 $2.00 $5.00
EXPLANATION:
At zero cost, quality is determined by sender's intrinsic motivation only.
Most senders are not intrinsically motivated to craft quality messages.
As cost increases, only messages the sender believes are valuable get sent.
This belief correlates (imperfectly) with actual value to recipient.
The cost functions as a filter: it doesn't make bad senders write better,
it makes bad senders stop sending.
I suspect that sealed messages would be, on average, much better than normal cold outreach. The incentive structure would make them behave differently. The cost functions as a forcing mechanism for quality.
Constraints paradoxically increase value. A sonnet isn’t worse than free verse because it has formal requirements. A haiku isn’t limited by its syllable count. The constraint is part of what makes it work. Similarly, a sealed message that required investment to send might be more valuable precisely because of that requirement, not despite it.
The trust layer we’re missing
We’ve built an internet where identity is cheap and attention is expensive, which is exactly backwards from how human relationships have worked for most of history. In the physical world, establishing identity took effort: you had to show up, be seen, accumulate a reputation over time. But once you’d established identity, you earned access to people’s attention by virtue of that identity.
Online, anyone can create an identity instantly. Email addresses cost nothing. Domains cost nearly nothing. You can be anyone. The result is that identity carries no information about trustworthiness, which means strangers have no reason to give you their attention, which means we retreat into closed networks of people we already know, which fragments the public sphere and makes it ever harder to make new connections.
The identity-attention inversion
HISTORICAL VS. DIGITAL IDENTITY-ATTENTION DYNAMICS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
PHYSICAL WORLD (pre-internet):
Identity Cost Attention Access
────────────────────────────────────────────────────────────────────
HIGH EARNED THROUGH IDENTITY
• Years to build reputation • Known actors get heard
• Physical presence required • Reputation opens doors
• Community vouching • Strangers can earn trust
• Credentials verified slowly • New connections possible
┌──────────┐ ┌──────────┐
│ Identity │ ─────── earns ──────────▶ │ Attention│
│ (hard) │ │ (given) │
└──────────┘ └──────────┘
DIGITAL WORLD (current):
Identity Cost Attention Access
────────────────────────────────────────────────────────────────────
ZERO HOARDED / DEFENDED
• New email in seconds • Everyone guarded
• Domains cost $12 • Filters block strangers
• No verification needed • Retreat to known networks
• Pseudonyms unlimited • New connections very hard
┌──────────┐ ┌──────────┐
│ Identity │ ─────── worthless ──────▶ │ Attention│
│ (free) │ │ (scarce) │
└──────────┘ └──────────┘
SEALED MAIL SYSTEM (proposed):
Identity Cost Attention Access
────────────────────────────────────────────────────────────────────
MEANINGFUL PURCHASABLE + EARNABLE
• Bond required • Strangers can pay to reach
• Attestation needed • Good actors build reputation
• Reputation accumulated • Bad actors priced out
• History matters • New connections possible again
┌──────────┐ ┌──────────┐ ┌──────────┐
│ Identity │ ───▶ │ Seal │ ──────▶ │ Attention│
│ (costs) │ │ (tracks) │ │ (priced) │
└──────────┘ └──────────┘ └──────────┘
Your seal would represent a history, a stake, a willingness to invest in communication. It wouldn’t guarantee that your message is valuable, but it would guarantee that you think it’s valuable, that you’ve bet something on that belief. That’s information recipients can actually use.
Implementation?
Phase 1: The bootstrap problem
ADOPTION CURVE CHALLENGES
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
THE CHICKEN-AND-EGG:
Senders won't get seals ◀────────────▶ Recipients won't publish
until recipients use it thresholds until senders
have seals
This is the classic two-sided market problem.
POTENTIAL BOOTSTRAP STRATEGIES:
Strategy 1: Premium Recipients First
─────────────────────────────────────────────────────────────────────
Target recipients whose attention is obviously valuable:
• VCs (flooded with pitches)
• Journalists (flooded with PR)
• Executives (flooded with sales)
• Creators (flooded with collabs)
These people would pay for a working filter.
Once they adopt, senders must follow.
┌─────────────────────────────────────────────────────────────────┐
│ Year 1: 10,000 high-value recipients │
│ Year 2: 100,000 senders get seals to reach them │
│ Year 3: Other recipients see value, adopt │
│ Year 4: Network effects take over │
└─────────────────────────────────────────────────────────────────┘
Strategy 2: Vertical Integration
─────────────────────────────────────────────────────────────────────
Build for one industry where both sides benefit:
• Recruiting: candidates pay to reach recruiters, recruiters pay
to reach candidates, both tired of noise
• Investor relations: founders pay to reach investors, investors
get signal about serious founders
• Enterprise sales: buyers get compensated for attention, sellers
get better targeting
Prove the model in one vertical, then expand.
Strategy 3: Integration with Existing Platforms
─────────────────────────────────────────────────────────────────────
Partner with:
• LinkedIn (already has InMail credits, imperfect version of this)
• Gmail/Outlook (offer as premium feature)
• CRM platforms (Salesforce, HubSpot integrate seal sending)
Leverage existing user bases rather than building from scratch.
Phase 2: Technical implementation
TECHNICAL ARCHITECTURE OVERVIEW
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
LAYER 1: PROTOCOL
─────────────────────────────────────────────────────────────────────────────
Extension to existing email standards:
New headers:
┌─────────────────────────────────────────────────────────────────┐
│ X-Seal-ID: a]b7c9d2e4f6a8b0c2d4e6f8 │
│ X-Seal-Attestation: <cryptographic proof> │
│ X-Seal-Payment-Ref: pay_1234567890 │
│ X-Seal-Reputation: 78 │
│ X-Seal-Tier: 2 │
└─────────────────────────────────────────────────────────────────┘
Recipients' mail servers verify seal before delivery.
Invalid/missing seals go to normal inbox (or spam).
Valid seals go to priority "Sealed" inbox.
LAYER 2: ATTESTATION SERVICE
─────────────────────────────────────────────────────────────────────────────
┌──────────────┐ ┌──────────────┐ ┌──────────────┐
│ │ │ │ │ │
│ Sender │────────▶│ Attestation │────────▶│ Recipient │
│ Domain │ │ Authority │ │ Server │
│ │ │ │ │ │
└──────────────┘ └──────────────┘ └──────────────┘
│ │ │
│ 1. Request seal │ │
│ (submit bond, │ │
│ verify identity) │ │
│ │ │
│◀───────────────────── │ │
│ 2. Issue seal │ │
│ (cryptographic │ │
│ credential) │ │
│ │ │
│─────────────────────────────────────────────────▶
│ 3. Send sealed message │
│ (include seal proof) │
│ │ │
│ │◀───────────────────────│
│ │ 4. Verify seal │
│ │ (check attestation, │
│ │ reputation, payment)│
│ │ │
│ │───────────────────────▶│
│ │ 5. Confirm valid │
│ │ (deliver to sealed │
│ │ inbox) │
LAYER 3: PAYMENT RAILS
─────────────────────────────────────────────────────────────────────────────
Options:
• Traditional: Stripe/payment processor holds funds, settles daily
• Crypto: Stablecoin escrow, instant settlement, programmable slashing
• Hybrid: Fiat on-ramp, crypto settlement layer
Key requirement: Payments must be:
├── Instant (or near-instant) verification
├── Reversible for abuse (slashing mechanism)
├── Low transaction costs (can't pay $0.30 fee on $1 message)
└── Internationally accessible
Why this probably won’t happen (and why it might)
BARRIERS TO ADOPTION
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
BARRIER SEVERITY NOTES
─────────────────────────────────────────────────────────────────────────────
Coordination problem ████████ Two-sided market, need both
senders and recipients
Incumbent resistance ███████░ Google/Microsoft have no
incentive to change
Philosophical opposition ██████░░ "Communication should be free"
is deeply held belief
Technical complexity █████░░░ New protocol, payment rails,
attestation infrastructure
Regulatory uncertainty ████░░░░ Payment regulations vary by
jurisdiction
User experience friction ████░░░░ Any new system is harder than
current email
International adoption ███░░░░░ Different countries, different
payment systems, different laws
COUNTERVAILING FORCES (why it might happen anyway):
Force Strength Notes
─────────────────────────────────────────────────────────────────────────────
Email is genuinely broken ████████ Everyone agrees current state
is terrible
Attention is valuable ████████ High-value recipients would
pay for working filter
AI making spam worse ████████ LLMs enable infinite
personalized spam at zero cost
Existing partial solutions ██████░░ LinkedIn InMail, Superhuman
priority, etc. show demand
Crypto infrastructure █████░░░ Payment rails for micropayments
now exist
Privacy regulation █████░░░ GDPR etc. creating pressure
for consent-based communication
First, there’s the coordination problem. Email works because it’s universal. Anyone can email anyone else, using any client, through any server. A sealed mail system would require widespread adoption to be useful, but it’s only useful once it’s widely adopted. This is the classic chicken-and-egg problem that kills most communication protocols.
Second, there’s the incumbency problem. The companies that dominate email (Google, Microsoft, Apple) have no particular incentive to change the system. They’ve built elaborate spam filtering precisely because the underlying economics are broken. Fixing the underlying economics would make their spam filtering less valuable, which isn’t a great pitch for getting them on board.
Third, there’s the philosophical resistance. Many people genuinely believe that communication should be free and frictionless, full stop. Any proposal that introduces cost feels like a step backward, a re-aristocratization of communication. I’ve tried to argue that some friction is actually good, that it creates the conditions for higher-quality communication, but I recognize this is a minority position.
But I can’t help noticing that the problems I’m describing keep getting worse. The volume of gray mail keeps increasing. The quality of cold outreach keeps declining. People’s email addresses become ever more carefully guarded, ever more hidden from strangers, because the cost of exposure is just too high. We’re moving toward a world where the only people who can reach you are people you already know, which is safe but also limiting.
I find it hard to believe that the current equilibrium - sending a message costs nothing and receiving messages costs everything - is stable in the long run. Something has to give. The alternative is that we’ll just continue retreating into smaller and smaller circles of pre-existing trust, letting the commons lie ruined, wondering why strangers never reach us with anything worth hearing.
The Penny Post was progress. But progress sometimes needs revision. Rowland Hill gave us the gift of friction-free communication, and we should be grateful. But two centuries later, we’re drowning in the abundance he made possible. Perhaps it’s time to take the stamp off the letter and put the seal back on.
Thanks for the post. You are describing a real problem, but one we have been living with for more than three decades, and it is perhaps too late to solve. I still use email socially, but some of my students recently told me that they didn't actually compose an email message until they started college. Younger people now use email mostly for some outer-circle professional correspondence and for technical reasons (passwords, codes, receipts, etc.). I expect we will see email losing what little relevance it still has as an actual communication platform between humans in the next few years.