Your board does not care about AI. They care about money, risk, and time.
That sounds blunt, but it is the key to getting approval. Every failed AI pitch makes the same mistake: too much technology, not enough business. This guide gives you the framework to present AI the way boards actually make decisions.
What Your Board Actually Wants to Know
Strip away the buzzwords. Your board is asking five questions:
- How much does it cost? Total investment — setup, tools, training, ongoing.
- How much will it save? Or earn. In pounds or dollars. Per month or per year.
- When do we see returns? Payback period. Not "eventually" — a number.
- What could go wrong? Risks. Data security. Regulatory concerns. Failure scenarios.
- What happens if we do nothing? The cost of inaction. Competitor pressure. Market trends.
Your entire presentation should answer these five questions. Nothing else.
The Framework: One Page, Five Sections
1. The Problem (2 minutes)
Start with the pain, not the solution. Describe the specific operational problem AI will solve. Use your own data:
- •"Our team spends 120 hours per month manually processing invoices"
- •"We lose an average of 3 deals per week because proposals take 5 days to prepare"
- •"Customer response time averages 4 hours — our competitors average 15 minutes"
Numbers from your business. Not industry averages. Your board knows when you are guessing.
2. The Solution (3 minutes)
Describe what AI will do — in plain English:
- •"AI will read incoming invoices, extract the data, match it to purchase orders, and route it for approval. Automatically."
- •"AI will generate first-draft proposals using our past wins, pricing data, and client history. Same day."
- •"AI will respond to routine customer enquiries instantly, 24/7, and escalate complex issues to our team."
No jargon. No "machine learning models" or "natural language processing." Just what it does and what changes.
3. The Numbers (5 minutes)
This is where you win or lose. Show three things:
Investment Required:
| Item | Cost |
|---|---|
| AI tool licences | $X/month |
| Setup and integration | $X one-time |
| Training (team time) | X hours |
| Ongoing maintenance | $X/month |
| Total Year 1 | $X |
| Saving | Monthly Value |
|---|---|
| Hours saved (X hours x $Y/hour) | $X/month |
| Revenue increase (faster proposals) | $X/month |
| Error reduction (fewer rework hours) | $X/month |
| Total Monthly Savings | $X/month |
Use our AI ROI Calculator to generate these numbers for your specific situation.
4. The Risks (3 minutes)
Address these before they are asked:
- •Data security: "All data stays within [provider]. Encrypted. GDPR-compliant. No data used for training."
- •Reliability: "We start with a pilot. AI handles the routine; our team handles exceptions. Nothing changes overnight."
- •Staff impact: "This replaces tasks, not people. Our team moves from data entry to analysis. From admin to advisory."
- •Vendor lock-in: "We use standard formats and APIs. If the tool does not work, we switch. No custom dependencies."
5. The Ask (2 minutes)
Be specific:
- •"We are requesting $X for a 90-day pilot focused on [one specific workflow]."
- •"Success criteria: [specific metric] improves by [specific percentage]."
- •"If the pilot succeeds, we present Phase 2 in Q[X]. If it does not, total exposure is $X."
Boards love pilots. Low risk, measurable outcomes, clear exit criteria.
Three Mistakes That Kill AI Proposals
Mistake 1: Leading with technology. Your board does not need to know how a neural network works. They need to know it saves $50,000 per year.
Mistake 2: Overpromising. "AI will transform everything" is not credible. "AI will cut invoice processing time by 70% in 90 days" is credible.
Mistake 3: No comparison to doing nothing. The alternative to investing in AI is not "stay the same." It is "fall behind while competitors adopt it." Make the cost of inaction explicit.
The One-Slide Summary
If your board only sees one slide, make it this:
Problem: [Specific operational bottleneck costing $X/year]
Solution: AI automates [specific task], saving [X hours/month]
Investment: $X over [X months]
Return: $X/year in savings + [qualitative benefits]
Payback: [X months]
Ask: $X for a 90-day pilot. Success = [specific metric].
What Happens After Approval
- Week 1-2: Select vendor, define scope, set up measurement baseline
- Week 3-6: Implement AI on chosen workflow, train team
- Week 7-10: Run with AI assistance, measure results
- Week 11-12: Compile results, present to board, propose Phase 2
The pilot proves the concept. The numbers prove the investment. The board approves the expansion.
Need help building your AI business case? Use our free ROI Calculator for instant numbers, or book a call to discuss your specific situation.