.webp)
AI is changing everything about physical operations, just as the steam engine, electricity, and the internet did when they first emerged. Many fear losing their relevance, but it's not AI that will replace people; rather, it's people who can use AI who will replace those who cannot.
People and businesses around the world are already leveraging AI to enhance operations and productivity. Cork Airport, for example, transitioned from reactive to proactive curbside management in just 20 days, saving 145 hours of congestion.
Toronto Pearson is capturing accurate queue-level wait time data, even as queue layouts constantly change which was considered impossible before. New Jersey Transit is capturing real-time ridership data as if they had a human data collector in every car, around the clock.
💡 Microsoft found that “for every $1 a company invests in AI, it realizes an average return of $3.5X, while the leading 5% of organizations realize $8X.”
Professionals and companies will lose competitiveness to those deploying AI and realize business values faster. What could the leaders do to stay relevant and successful in the era of AI?
Leaders are blocked by budgets
Unfortunately, many enterprise leaders who recognize the importance of fast AI adoption are blocked by budget constraints. Budgets set at the end of each fiscal year typically only include known expenses, often excluding new and emerging technologies like AI and its rapid breakthroughs.
It’s like trying to buy mechanical pencils when your budget is limited to pens and pencils. You know mechanical pencils are superior, but it’s hard to find the right item in the master budget to procure them.
Waiting years for the budget to become available for adopting AI is not an option, as the cost of delay is huge:
- Staying behind in a business environment which is becoming vastly more agile and missing out on the previous stated 3.5X to 8X returns using AI technology.
- Spending time on tedious tasks day-to-day when competitors have effectively automated them.
- Competing with AIs that are years ahead due to the vast data, experience, and improvements they've gained over time.
With the 2024 budget season right around the corner, it’s essential for leaders to secure the budget to stay competitive and relevant.
Why budget for AI Initiative
If your budget season isn’t over yet, establishing a new budget category specifically for AI can help fast-track adoption. Unlike legacy technologies and software, AI doesn’t fit neatly into traditional budget categories like software, facilities, and training. It often spans across these categories over short timelines.
Why? Legacy technologies and software are point solutions designed to handle one specific use case. You can easily categorize them into their respective budget items. It’s like an alarm clock that you can categorize under “time management,” but AI is like a smartphone that provides value way beyond time management.
How AI is evolving
Answering strategic questions like “How efficient are our operations at delivering an amazing customer experience, and where could we improve further?” required IT departments to procure dozens of different solutions, each capturing a part of business process, combine them, and try making sense of it all.
AI technology has advanced to the point where a single AI model can replace dozens of different solutions collecting data and the tedious process of making sense of them. Within weeks, AI can produce holistic operational insights. Business leaders can now have a live snapshot of an entire business process on one screen like the customer journey.
As a result, this single AI and data solution spans across line items used to procure legacy solutions but also budgets for data management, business intelligence, sensor technologies, and human capital among others. Instead of writing off bits and pieces from these items scattered across the master budget, having an independent budget for AI initiative streamlines budget allocation.
Measuring ROI for budget allocation
Establishing a separate budget category for AI not only facilitates procurement but also aids in calculating the return on AI investment. There are two components you should consider:
- Cost savings from replacing multiple legacy solutions collecting data with a single AI solution.
- The new business value created by having AI’s holistic insights compared to the individual pieces of data it replaces.
If your budget season is already over, you can still utilize the budgets set for legacy solutions and software to procure AI. It’s always recommended to start small and scale once you’ve verified the significant value of AI. This means you can use the budget set for a single use case or legacy solution initially, and expand to other items that AI can handle.
To calculate the ROI of your AI initiative without an independent budget, you can use a model to consolidate the costs you’ve saved and the business value you’ve produced from a unified AI and data solution. Use it to create a new budget category for next year.
AI budget deployment should be iterative
Establishing a new budget category for an AI initiative doesn’t mean rushing a full-scale AI deployment hastily. Rapid AI deployment for the sake of AI deployment brings risks of suboptimal results and potential failures, ultimately leading to the full abandonment of AI adoption. The cost of delaying AI adoption leads to losing a competitive edge, but it’s better than no adoption at all.
You can mitigate these risks with an iterative AI budget deployment - starting with one use case in a confined area and scaling up. This approach also allows you to validate if the single AI model you’ve procured or developed can actually scale across the entire business process for rapid adoption and greater business value.
During your first phase of AI deployment which can be done in days, you should look for:
- How quickly AI adapts to your unique business environment to capture data accurately (For reference, it took 8 days for Cork airport)
- Whether the AI is meeting its performance KPIs in the real-world environment, not just in the lab.
- Whether the data and insights that AI produces are relevant to your business.
Remember starting small doesn’t mean testing waters, but laying the groundwork for a system that can grow with your needs. 90% of enterprise initiatives fail to do this and end up investing in dead-ends not a springboard to broader application. Attend our 30-minute virtual session with Carnegie Mellon AI experts to avoid the common pitfalls that hinder many AI projects.
Understanding the cost of AI procurement
Unlike legacy solutions requiring a carefully outlined budgeting plan due to their capital-intensive, complicated deployment process, the cost structure of AI solutions is much simpler and straightforward.
Companies can leverage their existing infrastructure to test various AI solutions to find what works best for them, prove its value, and scale across the entire enterprise. With that in mind, here's a detailed breakdown of each cost component of AI.
Minimal CAPEX
Most modern AI solutions are cloud-based, meaning there’s often no need for significant upfront hardware investments. Choose AI solutions that leverage existing infrastructure to avoid waiting for next year’s budget and quickly roll out AI solutions to improve your bottom line.
Instead of CAPEX, consider setting up a separate line item just for introducing various AI solutions and validating their value. This provides a comfort zone for teams to learn about the new, rapidly evolving technology and to figure out what proves to be valuable, impactful, and meaningful. It also gets you into the mindset of iterative AI budget deployment: Start small, validate value, and deploy full-scale.
OPEX
There are one-time setup fees to integrate the AI solution into your existing systems, such as middleware, connectors, or custom development. Some AI vendors work closely with a variety of system integrators to make this process seamless and keep costs competitive.
Make sure to budget for ongoing costs for cloud-based AI platforms or tools. These can be monthly or annual and may vary based on usage or features accessed. If you’re building AI in-house, these fees can include data management (collecting, cleansing, and storing), employee salary and training, as well as AI model maintenance fees, which can be significantly higher than procuring an AI solution from vendors directly.
Determining the right budget size
Deciding on the appropriate budget for AI can be daunting, especially for companies taking their initial steps into this technological arena. Fortunately, a larger budget isn't a prerequisite for success in AI deployment. In fact, hefty investments can introduce unnecessary risks and may slow down the deployment process and the rate of end-user adoption.
To give you an idea of current investment trends, Omdia's survey of 369 companies revealed the following for AI budgets in 2023 consistent across companies of various sizes, regions, verticals, and industries:
- 42% allocated $1 million or more.
- 23% allocated $2 million or more.
- 9% allocated $5 million or more.
Companies with established AI applications, in particular, were observed to invest more substantially, indicating that businesses increase their AI budgets after validating the business value for the technology.
With AI technologies becoming more advanced and accessible, significant achievements are now possible without massive spending. For instance, a budget of $1 million provides a broad scope to experiment with various AI solutions and identify those that deliver tangible benefits to your business.
When planning your AI budget, focus on strategic impact rather than convenience:
- Avoid the trap of opting for the simplest AI solutions, such as chatbots, merely because they are widespread and straightforward to implement.
- Target AI solutions that are in line with your company's strategic goals and that bolster your competitive advantage.
- Employ a thorough AI buying guide to sift through the marketing noise and choose a solution that genuinely fits your business requirements.
By grasping the scalable nature of AI and starting with targeted, manageable deployments, your AI initiative is more likely to succeed, enhancing operational efficiency and significantly improving your bottom line.
