Home Impact Perspectives About

Cobalt Perspectives

From AI activity to underwritten EBITDA.

Practical frameworks for private-equity investors, operating partners and management teams seeking to identify AI-enabled value, convert it into executable EBITDA and cash, and sustain the result from diligence to exit.

Featured research

AI to EBITDA

Most portfolios do not lack AI activity. They lack a defensible path from use case to changed commercial behaviour, executable EBITDA and cash. These two briefs provide the investor and board view.

Executive brief

From AI opportunity to executable EBITDA

Six pages on value at stake, the commercial gap, the AI acceleration effect, Evidence Grade, the EBITDA-to-cash waterfall, and governance from diligence to exit.

Download the six-page brief →

Board brief

Five questions before approving the next AI initiative

A one-page provocation for testing ownership, changed decisions, executable economics, operating-model implications and exit relevance.

Download the board brief →

Relevant delivery experience

Pricing and margin recovery

Global outsourcing services provider

  • >$80M annualised margin improvement
  • 370 basis points of gross-margin expansion
  • Pricing architecture rebuilt across 12 service lines
See selected impact →

The five-part practitioner series

Each article examines a different part of the same problem: why functioning technology often fails to change commercial behaviour, and what investors and operators must redesign to turn it into value. Parts I and II establish the case. Parts III, IV and V provide the architecture, diagnostic and execution frameworks.

Full articles remain available on request. Each article stands alone, but the series is designed to build.

The Series

AI and the Operating Model Gap

Part I · The Case

AI and the Operating Model Gap

Why organisations are failing to capture the AI opportunity, and a practitioner's framework for closing the gap from diligence to exit.

AI initiatives rarely fail only because of the technology. They fail when decision rights, incentives, commercial ownership and operating rhythm remain unchanged. This article identifies five recurring failure modes and the structural fixes required to convert working tools into P&L impact, from diligence to exit.

"The tool worked. The system didn't. A faster broken process is still broken."
Request this article →
The Multi-Agent Operating System

Part II · The Case

The Multi-Agent Operating System

Why the next wave of AI value creation has nothing to do with prompts, and everything to do with architecture, orchestration, and agent-to-agent commerce.

The next wave of value creation will come from coordinated workflows, not isolated copilots. This article maps the move from point solutions to multi-agent operating systems and explains the data, orchestration, governance and accountability required to scale without creating a larger collection of expensive pilots.

"The orchestration layer connects technical capability to an operating model that can use it."
Request this article →
The Commercial AI Stack

Part III · The Work

The Commercial AI Stack

What it is built from, how the layers connect, which tools sit in each, and what it looks like deployed across pricing, sales, customer success, and operations.

Three layers (Data & Integration, Orchestration, and Commercial Applications) determine whether AI can scale across pricing, sales, customer success and operations. This article maps the stack, shows where value commonly leaks, and illustrates how commercial workflows must change for the technology to affect revenue or margin.

"Most organisations buy at the application layer before fixing the foundation beneath it."
Request this article →
The AI Value Diagnostic

Part IV · The Work

The AI Value Diagnostic

A scored self-assessment framework for identifying where your organisation actually stands, and where to focus first.

Before deploying the next tool, answer six questions. Is the data reliable? Is there a named commercial owner with a specific EBITDA number? Are outputs embedded in the decisions that change commercial behaviour? The diagnostic scores six dimensions, locates the organisation's current maturity, and produces a prioritised action list by business archetype.

"A plan without ownership is a wish list. The diagnostic identifies not just the technical gaps, but the accountability gaps."
Request this article →
The 100-Day Commercial AI Programme

Part V · The Work

The 100-Day Commercial AI Programme

A structured execution framework for moving from diagnostic to measurable commercial impact, for PE-backed businesses and corporate transformation leaders.

A structured 100-day programme turns diagnosis into evidence. Three phases (Diagnose, Sprint & Build, and Embed) connect use cases to decision rights, operating rhythm, P&L ownership and measurable outcomes. The article also covers five failure modes and starting points for transactional B2B, complex enterprise and recurring managed-services businesses.

"100 days is long enough to demonstrate real commercial impact and short enough to maintain organisational focus."
Request this article →

On LinkedIn

Ongoing thinking on value creation, transformation and AI deployment.

Follow Markus Schneider for shorter observations drawn from portfolio work, market research and the practical realities of delivering change in PE-backed businesses.

Follow on LinkedIn →

Where is AI supposed to move the P&L?

If a portfolio company has active pilots but no defensible value bridge, we can assess the commercial opportunity, execution constraints and evidence required for the next investment or board decision.