BCG X: It is about the people, going big, and capturing the value with a tailored solution

2023-05-09 · Source: Mike Talks AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, quick

Summary

Evan Shellshear, co-author of "Why Data Science Projects Fail" and a Principal at BCG X, discusses critical lessons for successful enterprise AI solutions. He emphasizes that 70% of project effort involves organizational buy-in and people management, 20% focuses on data acquisition and cleaning, and only 10% is dedicated to algorithms, a concept known as the 70-20-10 rule. Shellshear advocates for pursuing large-scale projects that demonstrate significant "value leakage" to maintain focus on substantial opportunities, citing an example where a $5M saving overshadowed a potential $30M gain. Furthermore, he argues that tailored AI and optimization solutions, rather than off-the-shelf products, are often necessary for truly transformative value, highlighting BCG X's "Build, Operate, and Transfer" model to ensure long-term success after consulting engagement.

Key takeaway

For CTOs or VPs of Engineering evaluating new AI initiatives, prioritize organizational change management and data strategy over algorithm selection. Your teams should focus on identifying and addressing significant "value leakage" opportunities to maximize ROI, even if it means a longer development cycle. Be prepared to invest in building tailored solutions and internal capabilities to operate and maintain them, rather than relying solely on off-the-shelf products for truly transformative outcomes.

Key insights

Successful enterprise AI projects prioritize people and data over algorithms, focusing on large-scale value.

Principles

• 70% of project effort is people-centric.
• Focus on "value leakage" for big impact.
• Tailored solutions capture more value.

Method

BCG X employs a "Build, Operate, and Transfer" model to ensure long-term solution viability and organizational capability development.

In practice

• Use the 70-20-10 rule for resource allocation.
• Quantify "value leakage" to justify large projects.
• Consider custom builds for transformative AI.

Topics

• Data Science Projects
• Enterprise AI Solutions
• Organizational Buy-in
• Build vs Buy
• AI Implementation

Best for: CTO, VP of Engineering/Data, Product Manager, Director of AI/ML, AI Product Manager, Executive

Related on AIssential

• Takeaways from CXO Insights: Exclusive Interviews with Top Operators — Effective data and AI initiatives require prioritizing decision context, repeatable governance, P&L impact, user adoption, and value-driven talent.
• From AI Experiments to Enterprise Value Driving Real Business ROI - with Dan Diasio of EY — AI transformation prioritizes workforce reinvention and new operating models over simple automation and headcount reduction.
• What If AI Is Just Simply The Latest Tech Evolution, Nothing More? — The AI market's rapid expansion demands disciplined investment and architectural intentionality, not just chasing hype.
• What CIOs Are Doing To Evolve Operating Models And Talent — Evolving IT operating models and talent is crucial for CIOs to translate strategy into measurable business outcomes.
• How Do We Fix This? — Focus AI investments on use cases where returns outpace costs to achieve economic viability and avoid negative unit economics.
• AI Can't Survive Your C-Suite’s Magic Thinking — Corporate "magic thinking" and business-driven data practices, not technical issues, cause 95% of AI initiative failures.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Mike Talks AI.