When it comes to AI budgeting, less is more, as long as you’re strategic

Whether we like it or not, there is incredible pressure on businesses to adopt AI. That pressure is only increasing as AI spending is expected to grow by nearly 30% over the next four years. Yet, questions remain on what and how much AI can actually do. As we head into budgeting season, this pressure leaves IT departments in every industry scratching their heads, trying to figure out how much they allocate to AI pilots. Spend too little, and the competition could develop an insurmountable advantage – spend too much, and you may find out that there isn’t enough ROI.

The reality is – AI needs a plan. With the right strategy, you can kick-start your projects by investing only a fraction of your budget. Well-executed AI initiatives have the potential to not only pay for themselves but also generate significant returns. Laying out a clear roadmap for adoption will put you ahead of the pack and position you as a leader in the AI game.

Let’s get this out of the way – spending more on AI doesn’t make you a leader in AI. Wedbush published a report that showed 1% of total revenue (approximately 8-10% of IT budgets) at major tech firms would be dedicated to AI. Meanwhile, a poll of CIOs revealed that 61% find it extremely challenging to prove ROI on their tech investments, with 42% not expecting positive ROI from AI projects within the next year.

Think about that for a moment – how much technology debt your IT department is already in because of projects showing no return or even negative returns. Now, you face the added pressure to bank on a new solution. The critical question is how to introduce AI within the constraints of your available budget while demonstrating ROI that drives widespread adoption and does not contribute to past failure.

The key is to start small – initiating pilot projects with a well-planned strategy rather than a comprehensive infrastructural overhaul. This approach allows your team the necessary time to transition and adapt while providing the flexibility for iterative improvements. You will have room for improvisation until you find a solution that works best for your business and see the impact in terms of revenue. By focusing on specific, manageable projects, you can gather valuable data and insights, refine your approach, and gradually scale up your AI initiatives.

How do you do it? Here are a few guiding principles that I believe will help you make the most out of your AI investment.

The first step is deliberately defining and choosing your first wave of AI pilots. There are people up and down your organization with a ‘great’ AI idea. If you chase all of them, and you’ll end up with a budget you can’t justify. On the other hand, if you only focus on the ideas that the loudest voices in the room suggest (or the voice with the highest title) you’ll likely miss out on the ideas most likely to have the biggest impact.

We often start with an AI Discovery process where all ideas are put on the table and each idea is objectively scored on a 12-point rubric based on technology fit, complexity, and ROI.

During a recent AI Discovery session with a CPA firm, the tax department wanted to automate tax return preparation, while the accounts receivable team sought to streamline payment processing. The objective behind both was to save more time. However, by scoring these ideas on project scale, complexity, potential ROI, strategic alignment, and feasibility, the firm decided to prioritize the accounts receivable project.

AI Discovery not only helps in selecting the most promising AI projects but also ensures that your initial investments are targeted and effective. It allows you to focus your resources on initiatives that are most likely to deliver measurable benefits, thereby increasing the likelihood of demonstrating a positive ROI.

Once you have conducted an AI Discovery and identified the most promising pilot projects, the next step is to focus on initiatives that are lower in complexity but offer a faster turnaround in terms of ROI. By starting with these types of AI pilots, you can create a self-funding transformation process that not only justifies the initial investment but also generates the momentum needed to fund subsequent projects.

Choosing AI projects that are simpler to implement and can quickly demonstrate tangible benefits allows you to build a strong business case for further investment. These initial successes can serve as proof points, showcasing the potential of AI to drive meaningful improvements and efficiencies within your organization. For instance, automating routine tasks or enhancing customer service through AI-driven chatbots are examples of low-complexity projects that can deliver quick wins.

Consider the case of the CPA firm I mentioned. Through AI Discovery, we identified automating their accounts receivable process as an effective initial project, quickly boosting their finances. This success enabled them to automate client document requests, saving 15% in tax return preparation time (which was the initial project suggested by the tax team).

Much like the CPA firm, you can invest the revenue or cost savings generated from pilot projects into more complex and ambitious AI initiatives. This approach minimizes financial risk and ensures that each step in your AI journey is backed by demonstrated success. It also helps in gaining confidence and buy-in from stakeholders, as they can see the direct impact of AI on the organization’s bottom line.

While focusing on launching pilot projects, you shouldn’t lose sight of the larger picture. The biggest ROI will ultimately come from company-wide AI initiatives. To achieve this, your entire organization needs to be AI-ready. This means fostering a culture that is open to innovation, investing in training and development, and ensuring that your infrastructure can support broader AI applications.

Moreover, you must keep in mind that not all pilot projects will succeed. Therefore, you should never put all your AI eggs in one basket. Diversifying your AI investments across multiple departments —such as finance, customer service, and operations— you can gather a wide range of insights that allows you to mitigate risks and identify the most impactful use cases.

Starting multiple AI pilots across departments will also accelerate your organization’s readiness for a complete AI transformation. Each successful pilot not only delivers immediate benefits but also builds the foundation for larger, more integrated AI projects. This approach ensures that your organization is continuously learning and adapting, making it more agile and better prepared for future challenges.