Advertising Process

How advertising flows from intent to action. The programmatic workflow and how DePIN data integration changes it.

Tight Five Job

Process is the system. It shows how advertising work moves from intent to action and back into learning.

System Step	Job	Good Sign	Bad Sign
Intent	Define the buyer, offer, and outcome.	The brief names the decision to change.	The campaign starts from channel preference.
Target	Build an audience from trusted signals.	Segments link to real behavior or context.	Targeting uses stale or guessed data.
Bid	Allocate budget at decision speed.	AI rules follow clear thresholds.	Automation optimizes a bad objective.
Deliver	Serve creative in the right context.	Message, medium, and moment align.	Frequency rises while relevance falls.
Measure	Compare reality with expectation.	MMM, MTA, and incrementality calibrate.	One attribution view becomes the truth.
Optimize	Feed learning into the next cycle.	Each campaign improves the next.	Reporting closes with no changed action.

The Advertising Flow

INTENT → TARGET → BID → DELIVER → MEASURE → OPTIMIZE
   ↓        ↓       ↓       ↓         ↓          ↓
Campaign  Audience  RTB   Creative  Attribution  Feedback
brief     build     DSP   serving   analysis     loop

Each step has protocols: automated rules that govern millisecond decisions.

Programmatic Workflow

How $168B in US digital ad spend flows through automated systems.

The Real-Time Bidding Cycle

Step	Time	Action	Technology
1	0ms	User loads page	Browser/app
2	10ms	Ad request sent	SSP
3	50ms	Audience matched	DMP/CDP
4	80ms	Bid calculated	DSP + AI
5	100ms	Auction resolved	Exchange
6	120ms	Creative served	Ad server
7	~seconds	User sees ad	Browser
8	~variable	User acts (or not)	Conversion tracking

100 milliseconds. That's the time from ad request to bid decision. Every step is automated. AI optimization at each stage compounds into massive efficiency gains.

Data Integration Protocols

First-Party Data Flow

User Action → Collect → CDP → Segment → Activate → DSP → Serve
                ↓                                           ↓
          Privacy consent                            Attribution

The shift: Third-party cookies → first-party data + AI modeling. Companies that build first-party data pipelines gain structural advantage.

DePIN Data Integration

How decentralized data sources feed the advertising pipeline:

DePIN Source	Data Type	Advertising Use	Integration Point
GEODNET	Precision location	Centimeter-accurate geotargeting	CDP/DSP audience layer
WeatherXM	Hyperlocal weather	Context-based creative triggers	DCO creative engine
Hivemapper	Fresh map imagery	Location intelligence, footfall	Analytics and planning
Helium	Connectivity data	Device/location intelligence	Audience enrichment

The Data Quality Protocol

DePIN Device → On-chain attestation → Verified data → AI processing → Audience signal
                       ↓
              Cryptographic proof of:
              - When collected
              - Where collected
              - How collected
              - Device identity

Why this matters: Advertising built on unverified data produces unverified results. DePIN attestations create a provenance chain from physical sensor to targeting decision.

Channel Protocols

Search (SEM)

Component	Protocol	Optimization
Keywords	Auction-based bidding	Quality score maximization
Creative	Text + extensions	CTR optimization
Landing	Conversion-optimized	CVR improvement
Bidding	AI-automated	Target CPA/ROAS

Social (Meta, TikTok, LinkedIn)

Component	Protocol	Optimization
Audience	Interest + behavioral	Lookalike expansion
Creative	Video + carousel	Engagement rate
Placement	Automated across surfaces	CPM efficiency
Attribution	Platform-reported	Independent verification needed

CTV (Connected TV)

Fastest-growing channel. $34.49B spend projected in 2025.

Component	Protocol	Optimization
Audience	ACR + identity graph	Household targeting
Buying	Programmatic guaranteed + RTB	CPM efficiency
Measurement	Incrementality + brand lift	Cross-device attribution
Format	15s/30s + shoppable	Direct response + brand

Programmatic DOOH

Component	Protocol	Optimization
Targeting	Geofenced audiences	Location + time triggers
Buying	Programmatic SSP	Daypart optimization
Creative	Dynamic, weather/event triggered	Context relevance
Measurement	Foot traffic lift	Attribution via mobile IDs

Measurement Protocols

The Three-Model Approach

Model	Frequency	Purpose	Output
MMM	Quarterly	Budget allocation	Channel-level ROI
MTA	Daily	Campaign optimization	Touchpoint-level credit
Incrementality	Monthly	Causal validation	True lift measurement

Verification Protocol

Ad Served → Viewability check → Brand safety check → Fraud filter → Valid impression
                                                                          ↓
                                                                   Attribution credit

Value Chain

Stage	Traditional	AI + DePIN Era	Margin Shift
Data collection	Cookies, surveys	DePIN sensors, first-party	→ Data providers
Audience building	DMP + third-party	CDP + AI modeling	→ Intelligence layer
Media buying	Manual IO	Programmatic + AI bidding	→ Automated
Creative	Agency production	AI-generated, DCO	→ Platform tools
Measurement	Last-click	MMM + incrementality	→ Analytics

The Flywheel

Data Collection → Audience Intelligence → Ad Delivery → Conversion → Revenue → More Data
       ↑                                                                           ↓
       └──────────── Revenue funds better data collection and targeting ──────────┘

Better data → Better targeting → Higher conversion → More revenue → Better data

The Evidence Loop

CAMPAIGN → MEASURE → COMPARE (vs expectation) → DIAGNOSE → OPTIMIZE → NEXT CAMPAIGN

This is the VVFL applied to advertising. Every campaign generates data that improves the next.

Stage	Metric	Diagnosis If Below Target
Headline	CTR	Headline didn't select or intrigue
Landing	Bounce rate	Page didn't deliver on promise
Conversion	CVR	Friction, trust, or relevance issue
Revenue	ROAS	Wrong audience or wrong offer

Context

• Advertising Overview — The transformation thesis
• Platform — Tech stack architecture
• Performance — What to measure
• Players — Who operates at each layer
• AI Data Protocols — Data pipeline architecture

Links

• IAB OpenRTB Specification — Industry standard for real-time bidding
• Display Lumascape — The canonical AdTech ecosystem diagram
• Google Ads Auction — How Google's auction works

Questions

If the entire RTB cycle completes in 100 milliseconds, what human decisions still matter — and at what timescale do they operate?

• When DePIN attestations create cryptographic proof of "where, when, how" for every data point, which step in the programmatic workflow becomes unnecessary?
• The flywheel says "better data → better targeting → higher conversion → more revenue → better data" — where does this loop break for most advertisers?
• If the verification protocol filters 19% of desktop clicks as invalid, what does that imply about every metric built on click denominators?
• Which channel protocol (Search, Social, CTV, DOOH) is most resistant to AI automation — and is that resistance a moat or a lag?