Learning from Mistakes: How AI Can Turn Poor Business Decisions into Advantages
Learn how to transform business errors into lessons with AI from Sergey Tokarev.
Failure and setbacks are a natural part of both business and life. The real issue arises when the same mistakes lead to repeated failures. However, these errors can be transformed into valuable lessons, and AI offers powerful tools to make this transformation possible. Sergey Tokarev, an IT investor and co-founder of Roosh, shares three practical ways to leverage AI for this purpose.
“I created my first website at the age of 15. Back then, the only way to figure it out—just like learning to use a computer—was through trial and error. When I later ventured into entrepreneurship, I followed a similar path. For every idea that successfully grew into a business, there were countless others that remained as concepts or failed during market testing,” reflects Sergey Tokarev on his entrepreneurial career and explains how AI can ease these challenges nowadays.
Identifying Cause-and-Effect Relationships
According to the investor, one of AI’s greatest strengths lies in its ability to identify complex, non-linear connections. “For example, if there is a drop in sales, we might attribute it to decreased demand or pricing issues. AI, however, can uncover unexpected correlations—such as the influence of weather, emerging trends, local nuances, or geographical factors.”
This ability to detect subtle patterns is what powers Netflix’s recommendation engine, which keeps users engaged for hours. A similar approach can be applied to analyzing sales trends, production workflows, logistics, and much more.
Building an “Institutional Memory”
“While the first method focuses on immediate issues, this one involves learning from historical data,” explains Sergey Tokarev. “AI can thoroughly analyze your organization’s past successes and failures using defined parameters. It then identifies patterns and proposes strategies tailored to your context.”
UPS’s ORION platform can be such an example. It integrates AI into logistics operations, analyzing customer data, route maps, pickup schedules, and historical performance to optimize delivery routes. This has enabled UPS to cut travel distances by millions of kilometers annually.
Tuning Into Your Customers
This approach is particularly effective for B2C businesses. Customer interactions—whether through sales teams, chatbots, or reviews—contain valuable insights for identifying and addressing mistakes.
“You do not need customers to explicitly point out issues; in fact, it is better if they do not. AI can extract insights from all communication channels to help refine your business,” says the entrepreneur. “For example, if multiple users ask the same question within a short timeframe, it is likely worth addressing it on your website or social media. While humans can catch some of these patterns, AI can uncover countless others that might otherwise go unnoticed.”
A Balanced Approach
Despite its potential, AI is not a substitute for human oversight. Consider the case of Zillow’s real estate division, Zillow Offers, which relied solely on algorithms to predict housing prices. Without human intervention, the system failed to adapt to unforeseen events like COVID-19 and labor shortages, leading to massive losses, layoffs, and the division’s eventual shutdown.
Sergey Tokarev insists that AI offers incredible opportunities to learn from mistakes, but it is equally important to learn from the missteps of others.
“Striking a balance between AI-driven insights and human judgment ensures the best outcomes.”
