Information Intermediary Evolution

What Is an Information Intermediary?

A business whose value = connecting buyers to sellers of information. Examples:

• Google (search ads)
• Amazon (product discovery)
• Yelp (restaurant reviews)
• Zillow (real estate data)
• Glassdoor (job market data)

The Agentic Disruption

Traditional value chain: User → Google → Website → User finds what they need

Agentic value chain: Agent → Direct API/database → Agent has all info needed

Agents don't need to be "directed" to a website. They can:

1. Query the source directly
2. Aggregate across sources
3. Evaluate without attention limits
4. Execute transactions without intermediary capture

Key Dynamics

1. Discovery Value Erodes

• Agents find products/services directly via API
• Search engine ranking no longer matters
• "Organic search" becomes irrelevant for agents

2. Comparison Value Shifts

• Agents compare at scale (1000s of options)
• Human-friendly comparison sites become obsolete
• Machine-readable data (specs, prices) becomes valuable

3. Review/Reputation Changes

• Agents verify claims directly (see verify-not-trust)
• Star ratings designed for humans lose meaning
• Verification databases gain value

4. Ad Targeting Disappears

• Agents don't see ads
• Ad networks dependent on human attention face existential crisis
• New models: incentive-based, procurement-based

Historical Parallel: Payment Intermediaries

The disruption of info intermediaries mirrors payment intermediary disruption:

• Old: Checks, wire transfers, slow and expensive
• PayPal/Stripe era: Friction reduced, fees still significant
• Crypto/Tempo era: Microtransactions at near-zero cost

Info intermediaries are at the "PayPal moment" — but the resolution is more dramatic because the value they provided (curation, discovery, matching) is fundamentally automatable.

Mediavine/Raptive Case Study

Business model: Ad network for content publishers

• Human readers see ads → publishers earn RPM $22-54
• Value = attention capture

Agent era problem:

• Agents don't see ads
• Content value = factual accuracy + verifiable claims
• Publishers need new monetization models

CPM/CPC/CPS History

These ad pricing models were invented for human attention economics:

• CPM: 19th century print ads → Internet impressions
• CPC: Search engine era → pay per human click
• CPS: Affiliate marketing → pay per human purchase

Agent era: These models need complete reimagining.

Who Survives?

Info intermediaries who can:

1. Add verification value — Be the trusted verifier
2. Provide unique data — APIs agents can't self-serve
3. Enable transactions — Fulfill, not just discover
4. Offer exclusivity — Data not available elsewhere

Related

• agentic-commerce — Commerce implications
• verify-not-trust — Trust shift
• no-app-commerce — Direct agent purchase model
• payment-for-order-flow — Ad revenue model analysis