How to Build a Business Case for Enterprise AI Investment
Most AI business cases fail before the technology is ever evaluated. They're built around the wrong metrics, framed for the wrong audience, and measured against the wrong baseline. Here's how to construct one that actually gets funded — and delivers.
Key Takeaways
- BCG (2023): 74% of companies struggle to achieve and scale value from AI investment — most fail at the business case stage, not the technology stage.
- Cost-saving business cases typically overstate savings by 40–60% once transition costs, retraining, and change management are factored in.
- A board-ready AI business case has four components: problem definition, value model, risk and dependency analysis, and operating model requirements.
- Different stakeholders measure success differently — CFOs want payback period; COOs want operational risk models; boards want strategic positioning.
- The right baseline is the best non-AI alternative available, not the status quo. Comparing AI against doing nothing overstates the investment case.
Enterprise AI investment is growing fast. According to McKinsey's 2023 State of AI report, 55% of organisations have adopted AI in at least one business function, and AI-related spend continues to increase across sectors. Yet BCG's 2023 AI survey found that 74% of companies struggle to achieve and scale value from those investments. The gap is not a technology problem. It is a business case problem.
Most AI business cases are built backwards — starting with a technology capability and working backwards to a justification, rather than starting with a business problem and working forwards to a solution. The result is a case that looks compelling in a slide deck and falls apart under scrutiny.
The Three Common Failures
Failure 1: Cost savings as the primary case. Efficiency gains are the easiest AI outcomes to model and the hardest to defend in practice. Actual labour displacement is constrained by employment law, union agreements, and the organisational cost of change. A business case built on headcount reduction often overstates savings by 40–60% once transition costs and retraining are factored in.
Failure 2: Technology-first framing. Presenting an AI investment as "deploying a large language model" or "implementing an agentic system" signals to a board or investment committee that the case is led by technical interest rather than commercial value. The technology is a means. The business outcome is the case.
Failure 3: Missing the risk dimension. AI investments carry implementation risk, operational risk, regulatory risk, and reputational risk. A business case that does not model these is not a business case — it is an optimistic projection. Sophisticated investment committees will ask about risk regardless; a case that addresses it proactively is far more credible.
What a Strong AI Business Case Contains
Framework Reference
The Board-Ready AI Business Case — Four Components
A case missing any component will fail scrutiny from at least one key stakeholder
Problem Definition
A quantified statement of the business problem — not "improve customer service" but the specific cost and impact measurable today.
Value Model
How the AI capability addresses the problem across three value sources — each quantified with explicit assumptions.
Risk & Dependencies
What must be true for the value model to hold — data readiness, integration complexity, regulatory constraints, change management.
Operating Model Requirements
How the organisation must change to capture the value — new workflows, accountabilities, and capabilities.
What each stakeholder needs to see
CFO
Payback period & IRR
COO
Operational risk model
GC / CRO
Regulatory exposure
CEO / Board
Strategic positioning
Imagine Works AI investment framework. The right baseline is the best non-AI alternative available — not the status quo.
A board-ready AI business case has four components:
1. Problem definition. A precise statement of the business problem, quantified where possible. Not "we need to improve customer service" but "unresolved first-contact queries cost £X per year in escalation handling and drive a Y-point reduction in NPS."
2. Value model. A structured analysis of how the proposed AI capability addresses the problem. Value typically comes from three sources: cost reduction (efficiency), revenue enablement (growth), and risk reduction (compliance, quality, reliability). A strong case quantifies all three and is explicit about which assumptions drive each figure.
3. Risk and dependency analysis. A clear-eyed view of what has to be true for the value model to hold. Data readiness, integration complexity, change management requirements, regulatory constraints, and operational governance all belong here. This section is where weak cases are exposed — and where strong cases build trust.
4. Operating model requirements. How the organisation will need to change to capture the value. AI tools do not deliver value independently — they require new workflows, new accountabilities, and often new capabilities. A case that models the technology without modelling the operating model is incomplete.
The Metrics That Matter to Investment Committees
Different stakeholders weight different outcomes. Understanding the primary concern of your decision-making audience shapes how the case is framed:
- CFOs prioritise payback period and IRR. Model the cash flows, not just the headline savings.
- COOs prioritise operational risk. Show what failure looks like and how the system recovers.
- General Counsel / Chief Risk Officers prioritise regulatory exposure. Address data governance, liability, and compliance obligations directly.
- CEOs and Boards prioritise strategic positioning. Frame AI capability as a competitive moat, not just a cost line.
The Right Baseline
One of the most common business case errors is comparing AI investment against doing nothing. The realistic baseline is not "status quo" but "the best non-AI alternative available." If a manual process can be improved with better tooling, that improvement should be the baseline against which AI economics are measured. Anything else overstates the AI case.
A business case that survives this test — that demonstrates clear superior value over the realistic alternative — is a genuinely strong investment thesis.
Imagine Works builds AI investment frameworks and operating models for enterprise organisations. Get in touch to discuss your AI business case.
Related Service
AI Strategy & Operating Model
Designing the AI strategy, vision, and operating model that aligns your entire organisation — from the boardroom to the workflow layer.
Explore this serviceMore Insights
More on AI Strategy
AI Centre of Excellence: Should You Build One — and What Should It Do?
As enterprise AI programmes mature, a recurring question emerges: who owns AI across the organisation? The AI Centre of Excellence is the most common answer — and one of the most misunderstood and misimplemented structures in enterprise AI. Here is how to get it right.
Workforce Planning for AI: Roles, Reskilling, and the Human–AI Team
The question enterprise leaders most commonly ask about AI and workforce is: which jobs will be automated? This is the wrong question. The right question is: how does AI change the work — and what does that mean for the people doing it? Here is how to plan your workforce transition correctly.
AI Portfolio Management: How to Prioritise Which Use Cases to Fund
Most enterprise AI investment is made project by project. Each decision is rational in isolation and collectively they produce a fragmented, duplicated, ungoverned portfolio that does not compound into organisational capability. Here is how to manage AI investment as a portfolio instead.