Perseverance and data management are key to success of AI investments

When it comes to investments in AI, perseverance and structured scaling pay off. A report from Protiviti shows that early-stage adopters often struggle, with 36% reporting ROI below expectations. But for organizations that push through, the picture changes dramatically – 96% report ROI that meets or exceeds expectations.

With AI technologies growing in maturity, organizations in a wide range of industries are scrambling to invest in the right tools to see the greatest possible returns on investment. But without good quality data, investment into AI may not pay off – at least at first. The equation is simple: If you feed an AI model bad data, you get bad results (a concept some refer to as ‘garbage in, garbage out’).

The importance of data is clear from the results of the Protiviti survey. The consulting firm reported the highest confidence in their data capabilities are significantly more likely to express that the returns from their investments in AI have exceeded their expectations.

“AI doesn’t exceed expectations by accident – it does so on the foundation of trusted data,” says Peter Mottram, global leader and AI leader at Protiviti.

“Issues like bias and misinterpretation can erode trust in AI – making strong data governance essential. As organizations advance along the AI maturity continuum, the nature of their data challenges shifts – though some issues remain of concern at all levels.”

AI maturity stages

Organizations at different stages of AI maturity can expect different ROI when it comes to AI tools. Working through the initial stages tends can be difficult, but the hard work is often worth it. The challenges that organizations face are different depending on its maturity stage.

For those at more advanced stages, the main issues are data reliability, completeness, and accessibility. Organizations that find themselves at the higher end of AI maturity did not rank these issues as their top concerns, though challenges like siloed data and integration difficulties continue to impact organizations across all levels of maturity.

The report found that 70% of organizations in stage 5 of AI maturity are very confident in their ability to obtain, organize, and understand the data required to support their AI goals. Stage 5 is the final stage of AI maturity in which transformation is the defining feature. At stage 1, the initial stage, less than 10% reported being confident.

Differences between sectors

Perhaps it comes as no surprise that the survey results show organizations in the technology sector to be the most confident in managing AI data. A total of 77% of companies in technology said they are confident in their data.

Those in the tech sector have more confidence in their data because of a stronger culture of innovation, earlier AI adoption, streamlined digital infrastructure, and fewer regulatory hurdles. They clearly have a competitive advantage in this domain.

In fact, the further away we get from technology, the less confident organizations are in their AI data. Leaders in consumer packaged goods, retail, manufacturing and distribution, were the least confident. This highlights the need for a special focus on data, and perhaps the need to bring in talent that can enhance data quality.

Avoiding the ‘perfect data trap’

In launching AI initiatives, some companies may sit back and wait for the perfect data to arrive. This can be a major hurdle. The truth is that progress starts with imperfection.

“Don’t let the pursuit of perfect data stall your momentum. Start with what you have – AI can actually help improve data quality over time,” says Mottram.

“Imperfect data still holds value: Even messy or incomplete data can uncover patterns, inform decisions, and support early AI use cases. AI tools themselves can help identify inconsistencies, identify and fill gaps, and enhance data quality – making AI part of the solution as well as the goal.”

The bottom line is that, rather than obsessing over perfection, organizations can benefit from embracing imperfection as a starting point. Real breakthroughs often happen when AI teams learn from trial and error.

“Progress isn’t about waiting – it’s about moving forward, smartly and boldly,” adds Mottram.

While AI investment is a huge priority for so many businesses, many leaders do not see it as a silver bullet for customer experience, according to a previous study from Protiviti. Instead, many companies are prioritizing data management technologies as the more impactful long-term solution.