What's Happening?
A new study titled 'Show Your Work,' conducted by Metric Sherpa in partnership with Sprinklr, has identified an 'AI credibility gap' within U.S. customer experience (CX) operations. The study, which surveyed 807 U.S. marketing, service, and CX practitioners
in September 2026, found that while 92% of leaders believe their customer-facing AI met or exceeded expectations, only 27% could precisely quantify these results with records or reports. This indicates a significant disconnect between confidence in AI performance and the ability to provide concrete evidence of its impact. The research also highlighted that 85% of respondents reported improvements in metrics their organizations did not track before AI implementation, suggesting that many perceived gains are not based on pre-established benchmarks. Furthermore, a substantial number of organizations (23%) set no expectations before launching AI, yet 85% of this group still claimed success, pointing to a lack of rigorous measurement practices.
Why It's Important?
This 'AI credibility gap' has significant implications for U.S. businesses investing heavily in artificial intelligence for customer experience. Without precise quantification of AI's impact, companies risk misallocating resources, making uninformed strategic decisions, and failing to realize the full potential return on investment from their AI initiatives. The study suggests that many perceived benefits, such as cost savings from increased speed, are often based on thin evidence rather than verifiable data. This can lead to a lack of accountability and transparency in AI deployments, potentially hindering future innovation and adoption. For stakeholders, including investors and executives, the inability to demonstrate clear, measurable outcomes from AI investments could erode confidence and challenge the justification for continued funding. The findings underscore the need for more disciplined measurement practices to ensure AI truly delivers on its promises and contributes tangibly to business objectives.
What's Next?
The report outlines five disciplines for organizations to bridge the AI credibility gap, framed as questions: Outcome, Owner, Baseline, Comparison, and Review. It recommends starting with one customer-facing AI application and applying an eight-question scorecard to identify areas where measurement practices are lacking. For each 'No' answer, a 90-day plan is suggested to close the gap, beginning with assigning a clear owner. This structured approach aims to help companies establish clearer goals, baselines, and review processes for their AI initiatives. By implementing these disciplines, businesses can move towards more precise quantification of AI's impact, fostering greater accountability and enabling more informed decisions regarding scaling, modifying, or discontinuing AI programs. The emphasis on assigning an owner and setting expectations before launch is crucial for driving more rigorous measurement and ensuring AI deployments are tied to tangible business outcomes.
Beyond the Headlines
The 'AI credibility gap' extends beyond mere measurement issues, touching upon broader ethical and strategic considerations for U.S. businesses. The tendency to claim success without quantifiable proof can foster a culture of overconfidence in technology, potentially masking underlying inefficiencies or misalignments with business goals. This lack of rigorous evaluation could also lead to a 'black box' problem, where the mechanisms and true impact of AI are not fully understood, hindering continuous improvement and responsible AI development. Furthermore, the study highlights a potential disconnect between technological enthusiasm and practical business application, suggesting that many organizations are adopting AI without a clear strategic framework for its integration and assessment. Addressing this gap requires not just better metrics, but a fundamental shift in organizational culture towards data-driven decision-making and a more critical approach to technology adoption, ensuring that AI serves genuine business needs rather than being perceived as merely a 'neat toy.'













