Martech in metamorphosis

Scott Brinker and Frans Riemersma frame 2026 as a chrysalis moment for marketing technology: AI dissolves old production constraints and exposes harder problems of context, governance, orchestration, and agent-readable infrastructure.

15,505Products in the 2026 landscape.

0.79%Net growth, effectively plateauing.

29,000+MCP servers indexed after 18 months.

Content

89%

Data

75%

Management

72%

Commerce

49%

Social

49%

Ads

50%

Key metrics

Selected numeric signals from the PDF capture the plateau, churn, AI adoption, and governance gap.

15,505 martech productsThe 2026 martech landscape contains 15,505 products.

0.79% net growthThe landscape grew by 121 net products from 15,384 in 2025, effectively plateauing.

1,488 products addedNew entrants fell 40% from 2,489 in 2025 to 1,488 in 2026.

1,367 products removedProduct removals increased 13% from 1,211 in 2025 to 1,367 in 2026.

29,000+ MCP serversMCP registries indexed more than 29,000 unique servers, more than twice the martech landscape count in 18 months.

73% with GenAI policySurvey respondents reporting a formal generative AI policy rose from 52% in 2024 to 73% in 2026.

8% full governance confidenceSAS research cited in the report found only 8% of organizations fully confident in broader AI governance readiness.

63.1% publish AI-optimized contentStructured Q&A and schema markup were the most adopted AEO practice in the cited survey.

13.6% measure AI inclusionOnly 13.6% measured AI inclusion rate and agent-referred conversion.

Transformation thesis

The report's caterpillar/chrysalis/butterfly model spans control of the conversation, AI in marketing, software, roles, and marketing operations.

Martech is in metamorphosis

The report frames 2026 martech as a chrysalis: old forms dissolve while a structurally different industry assembles.

Control shifts to customer agents

The conversation moves from marketer-controlled funnels toward AI agents acting on buyers' behalf.

AI everywhere, integrated nowhere

Most organizations are in a necessary messy middle where every SaaS product has AI features but workflows remain fragmented.

Context-as-a-Service

The destination is platforms that deliver the right data, content, and capabilities to the right agent at the right moment.

Marketing ops becomes context engineering

Marketing operations shifts from system administration toward orchestrating data, content, tools, and instructions for agents.

Market in motion

Flat net growth masks churn: fewer entrants, more exits, and pressure on weakly differentiated torso vendors.

The stack is stratifying

The answer to consolidation versus fragmentation is neither; creation, orchestration, and autonomous action layers obey different competitive physics.

RAG is connective tissue

Many use cases reduce to retrieving the right proprietary context, generating accurately, controlling permissions, evaluating, and logging.

The middle layer becomes center of gravity

Decisioning and orchestration become the intelligence layer between foundational data and activation channels.

Where marketers are flying with AI

AI adoption increased across all six landscape categories in the 2026 survey of 208 marketing and operations leaders.

Advertising & Promotions

AI adoption rose from 30% in 2024 to 50% in 2026 (+20pp).

Content & Experience

AI adoption rose from 79% in 2024 to 89% in 2026 (+10pp).

Social & Relationships

AI adoption rose from 33% in 2024 to 49% in 2026 (+16pp).

Commerce & Sales

AI adoption rose from 28% in 2024 to 49% in 2026 (+21pp).

Data

AI adoption rose from 61% in 2024 to 75% in 2026 (+14pp).

Management

AI adoption rose from 58% in 2024 to 72% in 2026 (+14pp).

Governance gap

Production use cases are ahead of authenticity, lineage, privacy, and broader readiness controls.

Production outruns controls

AI copy production sits at 91% adoption while content authenticity and AI detection sits at 37%.

Readiness is thin

Only 8% report full confidence in broader AI governance readiness; policy is a start, not a finish line.

MCP and agent infrastructure

The report treats MCP as a shared protocol layer for a world where agents need tool and data access everywhere work happens.

29,000+ MCP servers

Independent registries indexed more than twice the martech landscape count in just 18 months.

Abundance creates context pressure

When connectivity becomes fluid, the harder question is what should be connected, governed, and surfaced.

Middle layer as center of gravity

The decisioning and orchestration layer becomes the intelligence layer between data foundations and activation channels.

Context

Agents need relevant customer history, stage, preferences, interactions, and profile data.

Constraints

Agents need eligibility rules, suppression rules, fatigue limits, preferences, and business rules.

Compromise and cognizance

Agents need arbitration across competing actions and closed-loop outcome memory.

FAQ

Each question and answer is a named RDF entity linked through the resolver.

What is the central metaphor of State of Martech 2026?

The report uses a chrysalis metaphor: martech is not simply evolving, but structurally transforming into a different organism.

What changed about the martech landscape count?

The landscape reached 15,505 products, up only 121 from 2025, for 0.79% net growth.

Why is flat growth misleading?

The flat headline hides major churn: 1,488 products were added while 1,367 were removed.

How does the report describe the AI-martech integration layer?

It says the ecosystem is converging on MCP and agent-facing connectors that let AI agents reach tools and data sources.

What does Context-as-a-Service mean here?

It means platforms deliver data, content, and capabilities to the right agent at the right moment with both deterministic reliability and probabilistic intelligence.

What is the main bottleneck after AI reduces production cost?

The bottleneck moves to relevance, context, governance, orchestration, and strategic coherence.

What is the build-versus-buy conclusion?

The report says build versus buy is the wrong question; leading teams often do both for the same use case.

How is the stack changing?

It is stratifying into creation, orchestration, and autonomous action layers rather than simply consolidating or fragmenting.

What is the governance gap?

High AI production adoption is not matched by detection, lineage, compliance, privacy, and broader governance readiness.

Why does RAG matter?

RAG exposes the common architecture underneath many use cases: retrieval, permissions, evaluation, human-in-the-loop, and logging.

What is the middle layer?

The middle layer is decisioning and orchestration between foundational data platforms and activation channels.

What are the four Cs for autonomous agents?

The middle layer supplies context, constraints, compromise, and cognizance so agents can act coherently and learn from outcomes.