By Brian Veloso, Managing Director at SAP Concur Canada
The AI boom has seen hundreds of billions in capital flowing into AI infrastructure, research, and business use cases. As AI adoption accelerates, Canadian and global organizations are increasingly focused on proving measurable return on investment (ROI).
In Canada, the challenge is particularly evident. Recent industry research shows that while 93 per cent of Canadian businessesare using AI in some form, only a small fraction is seeing measurable financial returns, with just 2 per cent reporting clear ROI from generative AI investments.
New data from the annual SAP Concur CFO Insights Survey explores how CEOs, finance leaders, and IT leaders measure and define the return on investment (ROI) taken from AI. It reveals several gaps in perception between teams deploying AI and the executives funding it.
Let’s discuss why today’s leaders can’t agree about AI ROI – and what they can do to get aligned.
AI ROI is real, but unsettled
According to SAP Concur research, leaders believe the infrastructure for AI success is largely in place. However, most of them agree that cross-functional collaboration, strong data foundations, and faster productivity gains can increase returns on AI. And as a result, 37 per cent of CEOs, 41 per cent of finance leaders, and 49 per cent of IT leaders say that AI ROI is meeting expectations.
While respondents are broadly satisfied with the direction of the AI journey, there’s hesitance among some of the leadership team. Around 38 per cent of finance leaders and 39 per cent CEOs report that it’s “too early to tell” if AI is delivering value. Meanwhile, IT leaders are the most likely group to report that it’s exceeding expectations (16 per cent).
These mixed views don’t signal failure, but they do point to a lack of consistent measurement frameworks across the business. IT oversees implementation and sees the operational wins firsthand, but this isn’t translating into a well-defined, transparent measure of ROI for leaders across the business.
Finance and business heads want to point to a specific, quantified outcome and be able to say, “AI delivered that result.” If they can’t, adoption could stall.
Leaders view AI through different lenses
The data highlights a misalignment in how AI is judged between departments. Over half (54 per cent) of the CEOs surveyed agree with finance chiefs (50 per cent) that difficulty evaluating ROI is halting adoption. IT leaders are consistently more positive, with the average respondent reporting more than three distinct factors increasing AI returns within the business. Finance leaders and CEOs report fewer than three on average.
Specifically, half of the IT audience reports high levels of adoption among teams can increase AI ROI, compared to 41 per cent of finance and 37 per cent of CEOs.
Respondents name a range of metrics when asked to consider which factors are most important in ROI evaluations. All leaders ranked productivity and time savings, accuracy and quality improvements, as well as cost savings as the top three factors for evaluating the ROI of AI initiatives. However, IT and finance departments in particular place a higher importance on risk and compliance, while CEOs expressed greater concern regarding technology vulnerabilities and data security.
Impact on customer experience is considered less of a priority, suggesting most AI initiatives are still internally focused. This internal focus may also explain why some organizations struggle to connect AI investments to direct commercial outcomes.
What’s the strategic sticking point?
Several blockers prevent AI from delivering higher returns. More than half (53 per cent) of finance leaders say the benefits are slow to appear within the organisation, while a similar percentage (51 per cent) admits that initial expectations were likely over-optimistic.
A faulty technical foundation could also block ROI from surfacing, as more than half of respondents (53 per cent) identify data quality and integration issues as a drag on performance.
CEOs and finance chiefs are responsible for delivering successful business investments that yield long-term value. For this audience, AI shortcomings can’t be dismissed as teething issues. Instead, they could mean pilots are never allowed to scale.
How to close the ROI divide
The building blocks for AI returns are visible, but they aren’t yet aligned. To prove value, leaders can take the following steps:
- Establish a unified framework for ROI
Different roles within the business should stop grading AI on different scales. Collaborate across the business to create a shared dashboard that tracks both “Hard ROI”, based on performance indicators like cost savings, time savings, and commercial uplift, and “Soft ROI”, measuring risk mitigation, talent deployment, accuracy or quality improvements. - Prioritise time-to-value in use case selection
The data shows that 54 per cent of leaders see slow returns. Address the stagnation by balancing long-term projects with quick-win deployments – such as automated expense categorisation – that can show measurable impact within a single quarter. - Solve the data debt
More than half (53 per cent) of leaders cite poor data quality as a challenge affecting the ROI killer. Organisations should treat clean, high-quality, integrated data as a prerequisite for any new AI implementation. No budget should be approved without a corresponding line item for data governance.ROI is here, but there’s work to do to ensure it builds. Today, finance, business, and IT leaders should agree on clear, measurable frameworks to assess the impact of AI investments. The future belongs to leaders who get the people who fund and deploy it within the organisation on the same page.
