Build a business case for the board

You’ve done the reading. You’ve seen what AI can do for businesses like yours. You’re fairly sure there are processes in your organisation that automation could improve significantly. But you’re not the final decision-maker – and the people who are aren’t convinced yet.

This is one of the most common situations we see. An operations manager, a marketing director, or a department head knows AI could help, but needs to get buy-in from a board or senior leadership team that’s cautious, sceptical, or simply unconvinced that now is the right time. This article is a practical guide to building the kind of business case that gets a ‘yes’ – or at least a ‘let’s try it’.

Sceptical boards respond to numbers. The most effective way to build urgency is to calculate what the current state of affairs is actually costing the business – in time, money, or missed opportunity.

• Time costs – how many hours per week does your team spend on the tasks AI would handle? Multiply by the hourly cost of those employees (salary plus overheads). If three team members each spend five hours a week on manual data entry, and the fully-loaded cost of their time is £25 per hour, that’s £19,500 a year spent on work a machine could do.
• Error costs – manual processes produce errors. What does a typical error cost when it reaches a customer? A wrong order shipped. A missed follow-up that loses a sale. Duplicate records that cause a billing issue. These aren’t hypothetical – your team can almost certainly name recent examples.
• Opportunity costs – what are your staff not doing because they’re tied up with manual work? What leads are going unresponded to because nobody’s available outside office hours? What campaigns aren’t running because the marketing team doesn’t have capacity? These are harder to quantify but often represent the biggest value.

Put these numbers on paper. Boards that are immune to enthusiasm are rarely immune to a clear financial picture.

A common reason boards say no to AI is that the proposal feels too big, too risky, or too vague. “Let’s implement AI across the business” is a difficult thing to approve. “Let’s spend £3,000 automating our lead follow-up process over the next four weeks and measure the impact” is a much easier decision.

Identify one specific, contained use case and propose it as a pilot. The ideal pilot project has a few characteristics: it addresses a measurable pain point, it can be built quickly (weeks, not months), it has a clear success metric, and it’s low-risk if it doesn’t work. Automating a manual admin process, deploying a chatbot on one channel, or connecting two disconnected systems are all good candidates.

A successful pilot does two things: it delivers immediate value, and it gives you evidence to support a larger investment. Boards that approved a cautious first step are far more likely to approve a second one when they can see the results.

Different boards care about different things. Before you present, understand what matters most to the people in the room.

• If they care about profitability – frame AI as a cost reduction. Show the time savings in pounds. Show how automation reduces the need for additional hires as the business grows.
• If they care about growth – frame AI as a capacity multiplier. Show how automation frees existing staff to focus on revenue-generating activities. Show how faster response times or better customer experience drives conversion.
• If they care about risk – frame AI as risk reduction. Show how automation eliminates manual errors, ensures compliance, and creates consistent processes. Show what happens when a key person is ill and they’re the only one who knows the manual process.
• If they care about competitiveness – frame AI as keeping pace. Your competitors are already doing this. Not in five years. Now. The question isn’t whether to adopt AI – it’s whether you can afford to be the last in your sector to do it.

Whichever angle you lead with, present the ROI simply. Cost of the project versus measurable savings or gains, with a realistic timeline for payback. For most SME AI projects, payback is measured in months, not years.

Experienced board presenters know that pre-empting objections is more effective than responding to them. Here are the most common ones and how to handle them.

• It’s too expensive – show the cost of the pilot, not the cost of a full transformation. Compare it to what the current manual process costs. For most SME-scale AI projects, the investment is measured in thousands, not tens of thousands.
• Our team won’t use it – propose AI that works alongside existing tools and processes, not instead of them. The best AI implementations are invisible to the end user – they automate the work nobody wanted to do in the first place.
• We don’t have the technical skills – you don’t need them. That’s what an implementation partner handles. The right partner builds, trains and maintains the AI tools – your team just uses them.
• What about data security – a valid concern and one worth taking seriously. Any credible AI partner will explain exactly how your data is handled, where it’s stored, and what safeguards are in place. Prepare this information in advance.
• Let’s wait and see – this is the hardest objection because it sounds reasonable. Counter it with the cost of delay – every month you wait is another month of manual costs, missed opportunities, and your competitors pulling further ahead.

Keep it short. One to two pages is enough for most boards. Structure it clearly: the problem, the proposed solution, the cost, the expected return, the timeline, and the risks. Include real numbers from your business, not generic industry statistics. And be honest about what you don’t know – boards trust people who acknowledge uncertainty more than people who promise certainty.

If you need help building the numbers, scoping the pilot, or understanding what’s realistic for your business, that’s something we can help with. Our discovery sessions are designed to give you exactly the kind of clear, practical information a board needs to make a decision.