In Part 1 of our Big Ideas 2026 coverage, we explored how machine legibility, agent-native infrastructure, and hyper-personalization are reshaping marketing.

Now a16z has released Part 2, and three predictions should stop every marketer in their tracks—because they’re not about marketing at all.

They’re about financial services, industrial infrastructure, and enterprise software. But the patterns they describe? They’re coming for MarTech next.

The Three Industry Shifts

1. The Electro-Industrial Stack

Ryan McEntush describes a fundamental shift in manufacturing—from software-first to systems-first. The companies winning aren’t just writing better code; they’re building complete ecosystems: supply chains, talent pipelines, and physical infrastructure.

The marketing parallel: The same shift is happening in MarTech. The winners won’t be point solutions (another email tool, another social scheduler). They’ll be comprehensive systems that handle end-to-end marketing operations.

2. Financial Services at the Tipping Point

Angela Strange makes the case that financial services and insurance are reaching an inflection point. Her argument:

Replacing legacy systems is now less risky than maintaining them.

For decades, financial institutions ran on COBOL mainframes because migration risk seemed too high. But now:

The talent that understood those systems is retiring
Lost revenue from outdated capabilities compounds annually
AI-native alternatives have matured enough to be viable

The result? Institutions that rebuild on AI-native infrastructure will unlock:

Unified data across previously siloed systems
Parallelized workflows that eliminate sequential bottlenecks
Expanded categories that legacy systems couldn’t support
Major margin gains from labor consumption by AI

3. The Dynamic Agent Layer

Sarah Wang predicts the most significant shift: traditional systems of record will lose their dominance as a new “dynamic agent layer” emerges.

Her example: IT Service Management (ITSM). Today, you file a ticket, wait for triage, wait for assignment, wait for resolution. The system of record tracks the process, but doesn’t do the work.

With a dynamic agent layer? You describe the problem, the agent resolves it immediately. The distance between intent and execution collapses.

The systems of record become the plumbing. The agents become the interface.

Why This Matters for Marketing

MarTech Is Financial Services, 10 Years Behind

Everything Angela Strange says about financial services applies to marketing technology:

Financial Services	MarTech
COBOL mainframes	Legacy marketing clouds
Siloed customer data	Fragmented channel data
Sequential approval workflows	Linear campaign processes
Compliance-driven paralysis	Brand-consistency paralysis
Massive margin opportunity	Massive efficiency opportunity

The MarTech stack is exactly where finserv was: bloated, fragmented, and increasingly risky to maintain.

Consider the average enterprise marketing stack:

6-10 different tools that don’t talk to each other
Customer data scattered across CRM, email platform, social tools, analytics
Workflows that require manual handoffs between systems
No single view of campaign performance across channels

The tipping point is coming. Just like in financial services, the risk of maintaining fragmented MarTech will exceed the risk of replacing it.

The Dynamic Agent Layer in Marketing

Sarah Wang’s “dynamic agent layer” is precisely what marketing needs.

Today’s marketing workflow:

Marketer identifies need (e.g., “promote our webinar”)
Creates brief in project management tool
Assigns to copywriter
Copy goes through review/approval
Approved copy goes to designer
Design goes through review/approval
Final assets go to channel specialist
Channel specialist schedules posts
Campaign goes live
Analyst pulls performance data
Report presented in next weekly meeting

This is system-of-record thinking: every step is tracked, routed, approved, logged.

Dynamic agent layer workflow:

Marketer says: “Promote our webinar across relevant channels”
Agent executes—immediately

The agent handles copy, design adaptation, channel selection, scheduling, and performance tracking. The marketer states intent; the agent delivers outcomes.

This isn’t science fiction. It’s the direction marketing is heading.

What Changes

From Tool Orchestration to Agent Delegation

Current AI marketing tools are still tool-shaped:

“Here’s an AI that helps write copy” (you still operate it)
“Here’s an AI that suggests send times” (you still approve it)
“Here’s an AI that generates images” (you still select them)

Dynamic agent marketing is agent-shaped:

“Handle our content marketing” (it operates autonomously)
“Optimize our email program” (it decides and executes)
“Find new acquisition channels” (it explores and tests)

The human role shifts from operator to director. You set goals and constraints; agents handle execution.

From Channel Specialists to Marketing Architects

When agents can execute across channels, the value shifts from “knowing how to operate the tools” to “knowing what outcomes to pursue.”

Channel specialists become less valuable. Marketing architects—people who understand strategy, positioning, and customer psychology—become more valuable.

The marketers who thrive will be those who can:

Define clear objectives for agents to pursue
Set appropriate constraints and guardrails
Evaluate outcomes and adjust strategy
Make judgment calls agents can’t make

From Periodic Campaigns to Continuous Optimization

Campaign thinking assumes human bottlenecks: we batch work because humans can only do so much.

Agent thinking is continuous: there’s no reason to batch when agents can execute in parallel, around the clock.

This shifts marketing from:

Quarterly campaigns → Continuous experimentation
Weekly reporting → Real-time adaptation
Annual planning → Dynamic resource allocation

The 10x Winners

Angela Strange mentions “10x winners” in financial services—companies that leverage AI to achieve dramatically better economics than legacy competitors.

The same will happen in marketing. Companies that rebuild their marketing operations on AI-native infrastructure will:

Move faster: campaigns that took weeks take hours
Test more: run experiments across channels simultaneously
Learn faster: unified data enables rapid optimization
Spend less: agent execution costs less than agency retainers or full-time specialists

The gap between AI-native marketing operations and traditional approaches will widen rapidly. Early adopters won’t just be marginally better—they’ll be categorically different.

The Opportunity in 2026

Strange ends with a call to founders: the opportunity is now. The same applies to marketing leaders.

The transition from fragmented MarTech stacks to AI-native marketing systems is beginning. The companies that make this transition early will compound their advantage.

This doesn’t mean ripping out every tool tomorrow. It means:

Audit your current stack: How fragmented is your data? How many manual handoffs exist?
Identify the highest-friction workflows: Where do campaigns get stuck waiting for humans?
Test agent-based approaches: Start with one channel, one campaign type
Build toward unified data: Every tool you add should contribute to a single customer view
Develop agent management skills: Learn to direct AI systems, not just operate tools

The dynamic agent layer is coming for marketing. The question is whether you’ll be riding it or competing against it.

For more on how AI is reshaping marketing operations, see our Big Ideas 2026 Part 1 analysis and The AI Results Gap.