Non-profit

AI costs and ROI for non-profits

The sticker price of an AI tool is not the cost of AI. For non-profits, where every pound is accountable to donors and trustees, understanding the real cost matters more than the marketing promise.

By Zakir Hoosen, Director, Fuzzelogic Solutions. Board-level guidance in plain English.

Boards ask me what AI costs. I tell them the same thing every time. The tool is the cheap part. The cost of AI is everything around the tool that no one put in the budget.

A non-profit that spends five hundred pounds a month on an AI platform and then discovers it needs a data cleanup costing ten thousand, a staff training programme costing five thousand, and an integration project costing fifteen thousand has not spent five hundred pounds. It has spent thirty thousand pounds in the first year, and the tool has not delivered a single measurable outcome yet.

This article is about the real cost of AI for non-profits, where the hidden expenses hide, and how to think about return on investment without the usual hype.

The visible cost and the hidden cost

Every AI tool has a visible cost. A subscription fee. A per-user charge. A setup cost. That number appears in the proposal and it looks reasonable. Most non-profit boards can absorb a few hundred pounds a month without changing anything else.

The hidden cost is everything else.

  1. The real cost of AI includes:
    • Data preparation and cleanup
    • Integration with existing systems
    • Staff time to learn and use the tool
    • Governance and oversight
    • Ongoing maintenance and updates
    • The cost of getting it wrong

The last one matters most for non-profits. If an AI system recommends the wrong beneficiary for a grant, the cost is not just the money. It is the trust. It is the relationship with the donor. It is the impact on the community you serve. Non-profits carry a reputational cost that commercial businesses do not.

What the research says

"Ninety percent of companies have launched some flavor of digital transformation, and only a third of the expected revenue benefits, on average, have been realized."

Source: McKinsey, Rewired to Outcompete

The two thirds that did not deliver did not fail because the technology was too expensive. They failed because the cost of preparation was underestimated. The data was not ready. The staff were not trained. The integration took longer than expected. By the time the real costs appeared, the budget was gone and the board had lost confidence.

"61% of CEOs say boards are rushing AI transformation, and around 40% of boards lack an informed view of how AI changes growth strategy."

Source: BCG, CEOs and Boards are aligned on AI in theory but divided in practice

Rushing past the cost stage is how boards approve something they cannot sustain. The first year looks fine because the subscription is cheap. The second year is when the maintenance, the updates, the retraining, and the data problems all land at once.

How non-profits should think about ROI

Return on investment for a non-profit is not just financial. It is measured in impact. If AI helps you reach more beneficiaries, that is a return. If it frees up staff time for direct service, that is a return. If it improves the accuracy of your reporting to donors, that is a return.

But those returns need to be real, not projected. The mistake I see most often is a proposal that promises to save forty percent of staff time on a particular task. That sounds impressive until you ask what staff time is being saved and what those staff will do instead. If the answer is they will take on more work, the return is not a saving. It is a redistribution.

For non-profits, the honest ROI calculation starts with three questions.

First, what is the problem? Not what the tool can do. What the organisation actually needs to solve. If you cannot name the problem in one sentence, you are not ready to buy the tool.

Second, what is it worth? If the problem is solved, what changes? Put a number on it if you can. If you cannot put a number on it, you cannot measure the return.

Third, what does it cost to get there? Not just the tool. Everything. Data, training, integration, governance, maintenance. Add it up and compare it to the value. If the cost exceeds the value, the answer is not yet.

The Isle of Man context

Non-profits on the Isle of Man operate in a specific financial environment. Funding comes from the Isle of Man government, from local donors, from UK charities, and from international sources. Each funder has different expectations around how money is spent and how outcomes are reported.

The Charities Registration Office in the Isle of Man requires clear reporting on how funds are used. AI can help with that reporting, but only if the data feeding the system is accurate and the system itself is governed. The cost of governance is not optional. It is part of the cost of AI.

The Isle of Man government has signalled that digital transformation is a priority for the island. That means more funding opportunities for non-profits that can demonstrate data competence. But it also means more scrutiny on how technology is used and whether it delivers what it promises.

The practical budget

If a non-profit is considering AI, the board should approve a budget that covers four things.

The tool itself. The subscription, the licence, the setup fee. That is the visible number.

The data. Whatever it takes to get the data ready. This is usually the largest single cost and the one most often left out of the proposal.

The people. Training, time, adjustment. Staff do not learn new tools overnight. Budget for the disruption.

The governance. Someone needs to oversee the system. That is a cost, whether it is a new role or an expanded existing one.

If the board approves a budget that only covers the tool, it is approving a project that will run out of money before it delivers results.

The honest version

Fuzzelogic is an Isle of Man firm that has spent nineteen years helping boards understand the real cost of technology change. We tell non-profit boards what most consultants will not. The honest answer is often that the cost of AI is more than the budget allows, and when that is the case, we put it in writing rather than build it anyway.

Your systems were built for a world before AI. Most can get there. We tell you which ones cannot.

Start with the assessment. Two to four weeks, fixed price, and you own the verdict and the roadmap whether or not we build any of it. When you are ready to talk AI, call Fuzzelogic Solutions and ask for Zak. www.FuzzelogicSolutions.com | info@FuzzelogicSolutions.com | +44 (0)1624 618950

Start with the assessment

Two to four weeks, fixed price, and you own the verdict and the roadmap whether or not we build any of it.

Get in touch

When you are ready to talk AI, call Fuzzelogic Solutions and ask for Zak.

www.FuzzelogicSolutions.com | info@FuzzelogicSolutions.com | +44 (0)1624 618950