To fully realize the possibilities of AI, organizations need to join IT for long games. A pursuit that requires patience, tenacity and strategic integrity. Quick wins are important, but they don't stand alone by providing meaningful value. Agile experiments are necessary, execution requires iteration, and initial challenges are inevitable.
Protiviti's first Global AI Pulse Survey highlights a compelling correlation between AI maturity and return on investment (ROI) and the disconnect between the expectations and performance of many organizations during the early stages of AI adoption. With over 1,000 respondents, the survey divides organizations from more than 12 industrial sectors into five stages of maturation.
-
Stage 1: Early – We recognize the possibilities of AI, but we recognize that there is no strategic initiative.
-
Stage 2: Experiment – Run small pilots to assess feasibility.
-
Stage 3: Defined -Integrate AI into business processes.
-
Stage 4: Optimization – Improve performance and scalability with data feedback.
-
Stage 5: Conversion -AI drives critical business transformation.
Expectations from AI Investment
As organizations progress through these stages, their satisfaction with AI investments will improve. In fact, of 50% of survey respondents who showed they were in the early stages of AI adoption (initial or experimental), approximately 26% reported that their AI return on investment was below expectations.
Of course, not all AI experimenters have experienced a decline in returns. In fact, the majority reported meeting ROI expectations, but the results showed concentrations that were slightly exceeded or slightly exceeded or significantly exceeded ROI expectations between mid-to-higher-stage groups of AI adoption.
In reviewing what distinguishes successful experimenters (experiments at the experimental stage of AI adoption exceeding ROI expectations), we found three persuasive attributes from those who did not.
-
It focuses on balanced key performance indicators (KPIs) and measures success by combining financial and operational metrics such as employee productivity, cost savings, and revenue growth.
-
Reporting challenges with skills and integration is because we tend to invest in training, uplifting, and collaboration beyond function.
-
Seek a variety of support, not just training, but also strategic planning support and data management tools.
Another: These successful experimenters emphasized their financial and operational outcomes more evenly, while others focused narrowly on cost savings.
Challenges facing AI experimenters
Many AI experimenters struggle not because of unrealistic expectations, but because of possible value where the goal is unclear or misunderstood. This challenge and difficulty in integrating AI into existing systems are two biggest hurdles that organizations face in the early stages of adoption (stages 1 and 2).
The integration issues peak at the mid-stage stages of AI adoption, but they start at the early stages. Interestingly, the challenges associated with understanding the most impactful use cases are the most serious at the earliest stage, immersed in the central stage, and resurfaced at the highest level of maturity, for a variety of reasons.
Of course, AI experimenters don't know how to apply AI strategically, so unlike more mature companies, technical compatibility remains a hurdle. Aggravating these issues is the fundamental issues of unclear or conflicting regulatory guidance, difficulty with data availability and access, and effective AI deployment.
A lack of structured approaches, unclear project objectives, and unreliable data often leads to overwhelming ROI in the early stages.
Redefine AI success
Another interesting finding from the study shows that as organizations move from stage 3 to 5, their success metrics evolve from cost reduction and process efficiency to revenue growth, customer satisfaction and innovation.
The good news is that organizations that start with an AI journey can revise courses by focusing on these success metrics. This starts with redefineing the success of AI. This means moving beyond short-term victory to sustainable change.
It's important from the start to understand clearly what you're trying to achieve with AI. They struggle to reach their full potential without clarifying what AI intends to achieve and how value is measured.
Early experimenters should try to build a solid foundation by:
Ask why? Why are they using AI? What specific problems are you solving?
Investing in data infrastructure is important. This step should include auditing existing data systems and implementing a robust data governance framework. Organizations are well suited to consider cloud-based platforms for scalability.
Early development of robust integrated strategies. Many existing systems were not originally designed to support AI. To overcome this shortage, organizations need to actively work on assessing and modernizing the infrastructure to handle AI workloads in the early stages. If data and business teams work together, they are likely to have greater success, with shared ownership of AI initiatives ensuring alignment and adoption.
Align your AI strategy with business goals and organizational culture: This is more than just a technical step. It involves ensuring organizational preparation and effectively managing cultural and operational changes.
Turn AI Trials into ROI Wins
This study is clear. ROI is very likely for early stage companies that can quickly test, learn and scale AI use cases. However, speed is important for capturing value, but it is important to recognize that AI experiments are ongoing and continuous iteration is required.
To win, think big, act quickly, and evolve continuously – never stop.
