Advertising Principles
The immutable truths. Channels change. Technology evolves. These don't.
The Five Principles
| # | Principle | Why Immutable | Implication |
|---|---|---|---|
| 1 | Attention is finite | 24 hours per day, fixed | Every ad competes with everything else |
| 2 | Relevance earns attention | Irrelevant messages get ignored | Data quality determines targeting quality |
| 3 | Measurement drives allocation | Budgets follow measurable returns | Attribution determines who gets funded |
| 4 | Trust decays with opacity | Unverifiable claims lose credibility | Transparent measurement wins long-term |
| 5 | Distribution follows audience | People move across platforms | Multi-channel or invisible |
1. Attention is Finite
Every human has the same 24 hours. Advertising competes not just with other ads but with sleep, work, relationships, and entertainment.
The math: Average person sees 4,000-10,000 ads per day. Recalls fewer than 10. The filtering is brutal and unconscious.
The implication: Interruption advertising has diminishing returns. The winners earn attention through relevance, not volume.
Ogilvy's truth: "The consumer isn't a moron; she is your wife." Respect attention or lose it.
2. Relevance Earns Attention
People don't hate advertising. They hate irrelevant advertising. The right message to the right person at the right time is a service, not an interruption.
The data connection: Relevance requires knowledge. Knowledge requires data. Better data from DePIN sensors and first-party collection enables precision that cookies never achieved.
Sutherland's insight: What people say they want and what they respond to are different things. Behavioral data reveals true preferences. Stated preferences reveal social performance.
The quality spectrum:
| Targeting Level | Data Source | Relevance | Cost |
|---|---|---|---|
| Broadcast | None | Low | Low per impression |
| Demographic | Census, surveys | Medium | Medium |
| Behavioral | First-party actions | High | Higher |
| Contextual | Real-time environment | High | Variable |
| Predictive | AI on multi-source | Highest | Highest |
3. Measurement Drives Allocation
What gets measured gets funded. What can't be measured gets cut first. This creates perverse incentives when measurement is wrong.
The attribution trap: Last-click attribution over-funds bottom-of-funnel and under-funds brand. Companies optimize for the measurable and starve the immeasurable.
The solution: Marketing Mix Modeling (strategy) + Multi-touch attribution (tactics) + Incrementality testing (truth). Three lenses, not one.
Blockchain opportunity: Immutable, timestamped ad delivery records create transparent attribution. No disputed impressions. No phantom clicks.
4. Trust Decays with Opacity
Ad fraud costs $84B+ annually. 19% of desktop clicks are invalid. When advertisers can't verify delivery, trust erodes and budgets shift.
The transparency thesis: Blockchain-verified ad delivery creates auditable proof chains. Every impression, every click, every conversion — recorded and verifiable.
The fraud problem:
| Channel | Invalid Traffic Rate |
|---|---|
| Desktop web | 19% of clicks |
| Mobile web | 9% of clicks |
| Mobile app | 22-28% of clicks |
| CTV | 18% of traffic |
The fix: On-chain verification. DePIN-powered location verification. Cryptographic attestation of real human attention.
5. Distribution Follows Audience
Audiences don't stay on one platform. They fragment across channels, devices, and contexts. Advertising must follow.
The channel shift: Linear TV → Streaming → Social → CTV → DOOH. Each migration creates a measurement gap that incumbents exploit.
The convergence: AI enables cross-channel optimization. One campaign, multiple channels, unified measurement. The future is omnichannel by default.
| Era | Primary Channel | Measurement | Winner |
|---|---|---|---|
| Broadcast (1950-2000) | TV, print, radio | GRPs, circulation | Big brands |
| Digital (2000-2020) | Search, social, display | Clicks, conversions | Google, Meta |
| AI + DePIN (2020+) | Cross-channel, CTV, DOOH | MMM + incrementality | Data owners |
The Ogilvy-Sutherland Framework
Two complementary lenses for advertising truth:
| Dimension | Ogilvy (Logic) | Sutherland (Psycho-Logic) |
|---|---|---|
| What works | Clear promise, specific proof | Counterintuitive truth, behavioral insight |
| Creative test | Does it sell? | Does it change behavior? |
| Measurement | Direct response, trackable | Behavior change, category reframe |
| Failure mode | Boring but clear | Clever but confusing |
Combined test: Does this ad make a specific promise (Ogilvy) that works with human nature rather than against it (Sutherland)?
The Test
Before any advertising investment:
| Question | Yes = Proceed | No = Reconsider |
|---|---|---|
| Does this respect finite attention? | Earns attention through relevance | Interrupts and irritates |
| Does this use quality data? | Behavioral, verified, specific | Guessed demographics |
| Does this measure outcomes? | Attribution is causal | Counting impressions |
| Does this verify delivery? | Transparent, auditable | Trust the platform |
| Does this follow the audience? | Cross-channel presence | Single-channel bet |
Minimum: Yes to 3 of 5.
Principles to Performance
| Principle | Performance Metric |
|---|---|
| Attention is finite | CTR, engagement rate, ad recall |
| Relevance earns attention | Conversion rate, ROAS |
| Measurement drives allocation | Attribution accuracy, incrementality |
| Trust decays with opacity | Invalid traffic rate, verification % |
| Distribution follows audience | Cross-channel reach, frequency distribution |
See Performance for the full metrics framework.
The Trust Problem
Who verifies the verifier? Advertisers don't trust publishers. Publishers don't trust exchanges. Agencies don't trust measurement vendors. This isn't a payment or compute problem — it's a coordination problem across execution, data, and verification. The meta-problem: the industry charges for trust but cannot prove it earned.
The Triangle
Advertising connects to AI Data and Robotics through the data flywheel:
AI DATA
/ \
(feeds) (trains)
/ \
ADVERTISING ——— ROBOTICS
(monetizes attention from robot-generated data)
- AI Data → Advertising: Better data enables precision targeting
- Advertising → AI Data: Ad revenue funds data collection infrastructure
- Robotics → Advertising: Robot-collected data creates new audience signals
Context
- Advertising Overview — The transformation thesis
- Data Flow — Naming system and data footprint
- AI Data Principles — Data truths that enable targeting
- Marketing — Business growth strategy
- First Principles — Broader principles framework
Links
- Ogilvy on Advertising — The original direct response discipline
- Alchemy — Rory Sutherland — Psycho-logic over logic-logic
- Antonio García Martínez — Substack — Former Facebook ads PM on identity and attribution
Questions
If the entire industry is a Rube Goldberg machine to answer one identity question, what happens when the user owns the answer?
- Which of the five principles breaks first when AI agents replace human attention — and does "finite attention" still apply to machine audiences?
- When Apple removed the identity join and cost Meta $10B, was that a privacy victory or a competitive weapon — and does the distinction matter?
- If trust decays with opacity, why do the most opaque platforms (Meta, Google) still capture the most budget?
- Does blockchain-verified delivery solve the trust problem, or does it just move the trust question from "was it delivered?" to "was it seen?"