Investing in AI and digitalisation: business case and financing options

AI applications and digitalisation projects are now firmly on the investment agenda for many SMEs, from smart scheduling software to automated quality control on the shop floor. The challenge is not so much finding financing, but the fact that many of these investments are intangible: there is no physical machinery or commercial property to pledge as collateral, and the returns lie in time savings or quality improvements that cannot always be immediately quantified in euros.

This article explains how to build a business case that convinces a lender, and which financing structures suit investments comprising primarily software, licences and implementation hours.

Why lenders can be cautious with digital investments

A physical asset such as machinery carries residual value that can serve as security, even if things go wrong. Software and AI implementations have virtually no residual value: a purchased licence holds little collateral value for a lender, and bespoke software even less so.

Consequently, a lender assesses this type of application predominantly on the quality of the business case and the company's cash flow, rather than on the collateral value of the amount invested. This means your justification must carry significantly more weight than with a traditional equipment investment.

Building a convincing business case

• Describe the current process and the bottleneck: where time, money, or quality are currently lost without the new technology.
• Quantify the expected savings or revenue as tangibly as possible, for instance in hours per week or error rates, and translate that into euros.
• Include the total investment, meaning not just licence fees but also implementation, training, and anticipated dips in productivity during the transition.
• Calculate the payback period based on realistic assumptions, and explicitly identify which assumptions carry uncertainty.
• Outline an alternative scenario: what happens to your competitive edge or cost base if you choose not to make this investment.

Worked example: payback period of an AI application

Consider this: a manufacturing business is evaluating an AI-driven system for visual quality control. The investment totals 85,000 euros, including implementation and training. The system replaces part of the manual inspection work, delivering an estimated saving of 2,200 euros per month in labour costs, plus a reduction in scrap and defects estimated at 800 euros per month.

That amounts to a combined saving of 3,000 euros per month, or 36,000 euros per year. Against an investment of 85,000 euros, the rough payback period is approximately 2.4 years. A finance provider will want to see how solid these assumptions are, for instance by seeing proof-of-concept data or pilot results from comparable applications.

Financing structures that suit digital investments

• A standard commercial loan, where the overall cash flow of the business serves as the underwriting basis rather than the asset itself.
• Deferred payment structures through the vendor or software provider, occasionally coupled with performance-linked milestones tied to achieved cost savings.
• Grants and tax relief schemes such as the WBSO for the development phase of an AI application, provided the project qualifies as research and development under RVO criteria.
• Innovation credit facilities and regional funding schemes, with up-to-date eligibility criteria available via RVO and Brainport Development for region-specific programmes.

Grants: valuable, but no substitute for a business case

Grants such as the WBSO can offset part of the development costs of an AI application, provided the initiative meets the statutory R&D criteria. The precise conditions, subsidy rates, and application windows are determined by RVO and change regularly; always check the current scheme parameters before factoring them into your budget.

A grant should complement your funding structure, not replace a robust business rationale. A lender will still want to know whether the project stands on its own merits and delivers a return, even without a subsidy safety net.

Brainport region: an ecosystem built for technology adoption

The Eindhoven area features a notably high density of enterprises actively driving technological innovation, reinforced by local knowledge institutions and proximity to the high-tech supply chain surrounding ASML. This has fostered an established regional network of advisors, research bodies, and Brainport Development to support business owners in structuring digital transformation projects and securing relevant schemes.

For an SME in the region, it pays to check with the Eindhoven municipality and Brainport Development to see whether active regional initiatives or collaborative consortia align with your specific digitalisation project, alongside national funding routes.

Next steps

• First build the business case, incorporating realistic assumptions and an alternative scenario if the investment is deferred.
• Investigate with RVO whether the project qualifies for WBSO or another innovation scheme.
• Speak with your bank and specialist finance providers to establish their lending appetite for intangible assets.
• Review your business case and funding options with an advisor before formally submitting an application.

Frequently asked questions

Can I fund an AI investment without physical collateral?

Yes, this is common practice, but the lender will place greater emphasis on the company's operating cash flow and the robustness of the business case, as the software asset itself holds minimal recovery value.

Does every AI application qualify for WBSO?

No, only projects satisfying the statutory research and development criteria set out by RVO. Implementing an off-the-shelf software solution typically does not qualify; bespoke, internally developed solutions often may.

How do I convince a lender when savings are difficult to quantify?

Use data from a pilot or trial phase wherever possible, and be transparent about which assumptions are firm and which carry uncertainty. A clear, prudent estimate is far more persuasive than an unsubstantiated optimistic forecast.