Get in Touch

You can win if you avoid these pitfalls.

 

We’re all witnessing the transformative potential of Artificial Intelligence (AI). From streamlining operations to unlocking new customer insights, the promise is immense. But as many business and IT leaders are discovering, turning those AI dreams into reality can be a bumpy ride.

How often do AI projects actually succeed? According to research by Harvard Business School’s Iavor Bojinov, the current success rate of AI projects hovers around a sobering 12%.

The current ROI on most AI initiatives is overwhelmingly low.

Leaders are compelled to invest based on AI’s ethereal promise, but in the short term, the current financial returns are just not there.

So, how can we buck this trend and ensure AI initiatives deliver tangible results? Bojinov’s insightful work, including his HBR article Keep Your AI Projects on Track, provides crucial guidance for leaders.

The Pitfalls and How to Avoid Them

Bojinov’s research highlights several common challenges that derail AI projects.

  • Lack of Clear Business Objectives: Many initiatives kick off with the “cool” factor of AI rather than a specific problem to solve or opportunity to seize.

To overcome this temptation, start with a clear business need. Don’t let the technology lead; let the business strategy drive your AI efforts.

Ask, What business outcomes are we trying to achieve? How will AI help us get there?

  • Data Dependencies and Quality Issues: AI thrives on data, but often the data is siloed, messy, or insufficient.

Leaders must understand the data landscape early on. Invest in data infrastructure and governance, as a big chunk of an AI project timeline is often devoted to data preparation.

Ask, Do we have the right data, in the right format, and of sufficient quality to train and deploy our AI models effectively?

  • Talent Gaps and Integration Challenges: Integrating these solutions into existing systems and workflows can be a major hurdle.

Building and deploying AI solutions requires specialized skills. Take a multidisciplinary team approach that bridges the gap between business understanding, data science expertise, and IT capabilities.

Ask, Do we have the right talent in place, or a clear plan to get it? How will our AI solutions integrate with our current technology stack and business processes?

  • Traditional IT Development Approaches: Long projects with a broad scope put AI projects at risk because they lock you into a solution from the start.

AI projects are often exploratory by nature. First, AI innovation is moving fast, so you’ll want to take the latest thinking into account. Second, you need to test and confirm what works for your company. Using an iterative or Agile approach, you can bring in new ideas midstream and “fail fast” to land on the best outcomes.

Ask, How can we create a nimble project? What mindsets and processes do we need to innovate, test new solutions, and incorporate lessons learned?

  • Overlooking the “Human” Element: Resistance to change, lack of trust in AI-driven insights, and inadequate training can all hinder adoption.

AI implementation isn’t just about technology; it’s about people. Bojinov underscores the need for change management and clear communication. For example, engage stakeholders early and often. Focus on how AI can augment human capabilities rather than replace them.

Ask, How will we prepare our employees for these changes? How will we build trust and ensure the adoption of AI-powered tools and insights?

The Takeaways for Leaders

Bojinov’s work provides actionable insights for successful AI initiatives:

  • Focus on business value first. Define specific, measurable business goals before diving into technology selection.
  • Prioritize data readiness. Invest in your data infrastructure and ensure data quality is a top priority.
  • Build collaborative, cross-functional teams. Foster communication and collaboration between business, IT, and data science teams.
  • Embrace iteration and agility. Adopt an agile approach that allows for learning and adjustments along the way. Break down large projects into smaller, manageable milestones.
  • Don’t forget the people. Plan for change management, communication, and training to ensure successful adoption.

The potential of AI is undeniable, but realizing that potential requires a strategic, business-driven approach. By learning from the challenges and adopting these principles we can improve the success rate of our AI initiatives and unlock the true value of this transformative technology.