As organizations increasingly adopt AI in hopes of accelerating growth and efficiency, many are discovering that rushing to see a return on investment (ROI) can lead to unexpected challenges. The desire for quick wins in AI adoption often results in poor planning and misplaced expectations, which can derail even the most promising projects. While AI holds great potential to transform business outcomes, realizing its full value takes time, careful strategy, and a measured approach.
According to Forrester, many companies that rush to demonstrate AI ROI may scale back prematurely, abandoning projects before they can fully mature. In their Q2 AI Pulse Survey, 2024, nearly half of AI decision-makers revealed they expect to see ROI from their AI investments within one to three years. However, another 44% acknowledged that the timeframe could be longer. The truth is, companies must be prepared for AI projects to take time to deliver measurable business value.
AI has become synonymous with productivity, efficiency, and innovation, but these benefits are rarely realized overnight. Companies often fall into the trap of expecting immediate ROI from AI investments, especially in industries where there is pressure to stay competitive. However, as AI and data science analyst Rowan Curran from Forrester points out, many organizations haven’t seen the anticipated returns, particularly when deploying general-purpose AI tools.
For example, generative AI systems like copilot programs may seem appealing on the surface, but their actual impact on productivity often falls short. These AI tools, while hyped for their potential, can struggle to demonstrate clear, quantifiable business outcomes. Organizations that focus too heavily on seeing quick ROI from these systems may find themselves scaling back too soon or abandoning projects altogether.
Curran stresses the importance of planning for long-term success rather than fixating on short-term gains. AI projects that focus on specific, targeted problems—like optimizing call center operations—tend to perform better in terms of measurable ROI. These projects demonstrate the potential of AI to transform business processes when deployed with a clear goal in mind and with time for refinement.
Measuring ROI from AI initiatives can be complex, particularly when organizations expect results too early. AI is not a one-size-fits-all solution, and its success depends on factors such as the nature of the project, the specific problem being addressed, and the quality of data being used.
For example, a call center that deploys an AI agent to assist with customer interactions may aim to reduce call times by 30 to 40 seconds per call. This is a measurable and trackable goal that can show tangible productivity gains over time. However, achieving the desired outcome is often a gradual process. An AI model might start with 75% accuracy and require months or even years of refinement, feedback, and iteration to reach 90% accuracy or beyond.
The same applies to many other AI projects, where ROI comes incrementally, rather than all at once. CIOs and IT leaders need to recognize that achieving ROI from AI is not a linear process and may require ongoing optimization before the full value is realized.
One of the driving forces behind poor AI planning is the fear of missing out (FOMO). Many organizations rush into AI adoption because they feel pressured by competitors or board members to keep up with the latest trends. However, this hurried approach often leads to failures, especially when organizations don’t take the time to fully understand how AI fits into their business model or what problem it is meant to solve.
Tony Fernandes, chief AI experience officer at HumanFocused.AI, emphasizes that many CIOs are forced into AI projects without conducting the necessary due diligence. This pressure can lead to unrealistic expectations and overstated progress, which ultimately results in disappointment and project abandonment.
Fernandes advises organizations to take a more measured approach—starting with methodical experimentation and gradually scaling as they identify proven use cases. Without proper preparation, organizations risk rushing headlong into dead-end projects that don’t align with their business needs or objectives.
Rather than jumping in with large-scale AI deployments, experts like Rob Owen, CIO of Sax, recommend starting small and experimenting with cost-effective AI solutions. Owen’s firm has successfully implemented AI in internal projects, such as enhancing help desk functions, by customizing and training AI models in-house. By starting with manageable projects, organizations can test the waters, gain insights, and iterate on their AI models without overcommitting resources.
Owen stresses the importance of “tinkering” with AI in affordable, low-risk ways before scaling up. He points out that it’s irrational to expect immediate ROI from AI investments. Most projects, particularly those involving AI, take 18 to 24 months to deliver returns. Therefore, patience and long-term thinking are crucial.
Deciding when to abandon an AI project can be one of the most challenging decisions for CIOs. The decision isn’t always clear-cut and depends heavily on the specific circumstances of the organization. If a project isn’t showing the desired results, leaders must evaluate whether the problem lies in execution, timing, or alignment with business goals.
However, organizations should avoid abandoning projects prematurely, especially if the project has shown incremental progress. Many AI initiatives require extended timelines to mature and demonstrate their full potential.
The rush to achieve quick ROI from AI can lead to costly mistakes and abandoned projects. Organizations must resist the temptation to seek immediate gains and instead take a thoughtful, measured approach to AI adoption. By setting realistic expectations, focusing on specific, targeted use cases, and allowing time for incremental progress, companies can unlock the true potential of AI without sacrificing long-term value.
In the end, AI success requires patience, careful planning, and a commitment to ongoing experimentation. Organizations that take the time to get AI right will be the ones that ultimately reap the most significant rewards.
