The physics of value

Most AI investment in marketing is going to the wrong part of the stack.

Most enterprise AI spend lands where value commoditises fastest. Value actually accumulates a layer below. Here is how to read your own stack, and how to stop the leak.

The Value Gravity stackValue evaporates as red particles off a thin top layer, gravity pulls it down through a bridge layer, and it accretes as white particles in a wide cobalt foundation.GRAVITY03EXPERIENCEEVAPORATES02ORCHESTRATIONTHE BRIDGE01FOUNDATIONACCRETES
Structure
03 layers, one flow
Force
Value accretes down
Failure mode
The Gravity Leak
i / The argument, in short

Activity is up. Capability is up. Value is not.

Every enterprise marketing function in Europe is running AI pilots. Few CMOs can point to a durable improvement in unit economics. The standard explanation is execution. The more precise one is structural: value in a marketing stack does not accumulate evenly. It pools in specific layers, and most current spend is going to the layer where value evaporates fastest.

When AI value cannot flow down into the foundation, it leaks. We call it a Gravity Leak.
ii / The model

Three layers. One directional force.

Economic value accretes downward through the stack. The deeper a layer sits, the more durable the value it holds.

The Value Gravity™ stack · value accretes downward ↓
03
Commoditises fast
Experience Layer
Content engines · Chat interfaces · Personalisation · Creative generation
The part buyers see. The part every vendor is shipping. The part that commoditises fastest.
02
The bridge
Orchestration Layer
Decision engines · Agentic workflows · Journey routing · Human-in-the-loop
Where AI output becomes governed enterprise action. The layer where value either lands or evaporates.
01
Value accretes here
Foundation Layer
Data model · Identity · Consent · Governance
Slow to build, expensive to change. Where economic value compounds over time.
Read the full essay →
iii / What it implies

So what?

I
Platform strategy

Context, not capability.

Platform wars have moved from model quality to context ownership. Adobe, Salesforce, and Microsoft are not competing on foundation models any more. The significant bets are at Layer 01: Real-Time CDP, Data Cloud, Microsoft Fabric. Switching costs, not capability.

II
ROI thinking

The maths is inverted.

Most AI ROI calculations are structurally inverted. The investments that excite the boardroom have the shortest value half-life. The quiet ones compound.

III
Pilot design

Structural, not executional.

The pilot graveyard is a structural problem, not an executional one. Most pilots test the tool, not whether value lands in the governed base. The fix is a different success criterion.

"Winning the AI era will require deeper integration into enterprise data, systems, and workflows."
Steve Lucas, former CEO of Marketo. In response to Arjen Segers on LinkedIn, 2026.
iv / Keep reading

This page is the invitation. The argument lives next door.

The full essay walks through the observation, the diagnosis, the model itself, the forces that produce it, the structural implications, and the practical questions it surfaces for any commercial leader trying to read their own stack.

Value Gravity™ is the framework. IDADAY International Services is the advisory practice that uses it.