Rebecca Surtay Explains How to Eliminate Bias in AI-Assisted B2B Projects
Updated: Jun 24
Written by: Henry Van Niekerk
More than 200 business solutions professionals from 20 countries, including the United States, China, the United Kingdom, South Korea, Singapore, Canada, Turkey, and India, took part in an exclusive masterclass hosted by Rebecca Surtay, a leading B2B marketing expert who has led large-scale projects for global corporations such as Samsung, Legrand, and Metro Cash & Carry. During the session, Surtay shared her expertise in adapting AI-driven forecasts to real-world business conditions.
The masterclass, titled "Why Perfect Forecasts Fail: Human Factors in AI-Assisted B2B Projects," was organized by the committee of the prestigious international NextWave Awards, which recognizes leaders who have made significant contributions to business development and the advancement of innovation.
Drawing on her extensive experience in B2B product development, Rebecca Surtay has developed a proprietary methodology that incorporates human factors into AI-assisted B2B projects. The methodology offers a new perspective on AI-assisted B2B projects, as even highly sophisticated organizations often rely heavily on digital models without critically assessing their practical feasibility.
Over more than a decade of working with leading technology companies, Rebecca Surtay has repeatedly witnessed cases in which AI forecasts proved inaccurate due to flawed problem framing and other human-related factors. During the masterclass, she presented a detailed analysis of a case in which Legrand Astana, a leading Kazakh provider of electrical infrastructure solutions, failed to secure a major contract despite an AI model assigning an 85% probability of success.
"The team fell victim to overconfidence bias. Participants viewed the project through the lens of previous successes and primarily relied on data that confirmed their assumptions," Surtay explained, noting that these cognitive biases are well documented in psychology.
Similar situations have occurred in other companies as well. Surtay cited the example of Samsung's implementation of a unified strategy based on standardized KPIs between 2014 and 2016. In three countries, the actual results fell 40% short of initial projections.
"At the time Samsung implemented the project, however, AI tools had not yet reached their current level of maturity," Surtay says. "Today, these tools are widely available, and their use increases the risk of errors stemming from bias in prompt design."
Surtay herself has extensive expertise in psychology and holds academic credentials in relationship management, as well as personal and professional development. As a result, she incorporates not only traditional B2B marketing practices into her work, but also advanced methodologies drawn from psychology.
During the masterclass, she introduced her proprietary debiasing techniques based on causal analysis and reference class forecasting. Surtay also demonstrated a protocol for validating AI-generated forecasts that involves both the project team and an independent external stakeholder.
"Companies are increasingly placing unquestioning trust in AI tools, relying on them to make major financial decisions and shape strategic business priorities. During the masterclass, Rebecca Surtay highlighted risks that may appear subtle at first glance, yet ignoring them can have devastating consequences for a company. What I found especially valuable was that the session not only identified these risks but also demonstrated how they can be mitigated. I believe the protocols developed by Rebecca Surtay should be considered industry standards for organizations that rely on AI in managerial decision-making," says Satenik Ghazaryan, Marketing Director of Armenia Digital Awards, an expert in marketing funnel development and a participant in the seminar.
