Lean & AI

1. Honest Assessment – What Kukoon Needs to Do Better

Paul opened with unusual transparency about Kukoon’s current challenges. They’re bringing back a second morning meeting in January because they discovered that reducing to one weekly meeting 18 months ago had failed to transfer lean knowledge to new team members.

The gap between experienced and new people has grown significantly, and without proper lean knowledge, people fall back into papering over cracks rather than identifying waste. Despite being seven years into lean, Paul admitted this was “pretty embarrassing” but necessary to acknowledge. They’re also implementing formal Kaizen targets for every department monthly, adding a lean metric to performance reviews, and creating a formal lean education structure. His willingness to share these struggles provided valuable lessons for everyone in the room about the ongoing nature of lean transformation.

2. Use ChatGPT for Everything – The Low-Hanging Fruit

Paul’s first recommendation was simple and emphatic: use ChatGPT at every single opportunity. Ask it every question, use it for analysis and comparison, and especially try it for things you think it couldn’t possibly help with. The learning process of discovering what AI can and cannot do is invaluable. He emphasized that ChatGPT has become phenomenal at teaching people how to be better prompters – the key is providing a feedback loop. When an answer isn’t right, tell ChatGPT why it’s not right, and it will help you understand what a better prompt looks like.

As part of personalization, ChatGPT encourages users to always help improve their prompting skills. This low-risk, high-value exploration costs $30 per month and provides immediate learning opportunities for everyone in the organization.

3. Build Automation Capabilities – Scripts, Coding, and Integration

Kukoon invested significantly in building a three-person business improvement team focused entirely on office-based improvements through data analysis and automation. They use a combination of Google Apps Scripts, Python scripts, and Microsoft Power Automate to create integrated automation systems. For example, they might set Power Automate to pull data from a specific channel, execute a Google Apps Script to manipulate that data into the required form, then return to Power Automate to send the processed file wherever needed.

Paul showed a revealing graph: over 11 months, their monthly hours invested in Power Automate (red line) started very high with practically no results, but the monthly hours saved (blue line) grew exponentially. By month 12, they were heading toward 150 hours saved monthly for less than 40 hours of input time. The key phrase Paul emphasized: “slow is smooth and smooth is fast.” The learning curve is difficult with minimal gains early on, but once you understand the tools, most Power Automates take less than an hour to write. Paul noted anecdotally that they multiply the time savings by three to account for the genuine value, including reduced defects, improved customer service, and eliminating the cognitive cost of task-switching.

4. Create Subject Matter Experts and Give Time to Learn

Within their business improvement team, Kukoon created specialists – one person focuses on Google Apps Scripts, another specializes in Python, and so on. This creates deep expertise where complex topics become owned by individuals who can deliver at high levels. Critically, Paul emphasized giving people time to learn and test, especially in this area where the learning curve is much steeper than traditional lean improvements.

The team became “prompting pros” because they invested time in understanding how to interact effectively with AI tools. Paul stressed that creating this specialization and allowing learning time was essential to moving beyond easy wins and low-hanging fruit toward maximum returns.

5. Focus on Maximum Returns – Not Just Easy Wins

Once Kukoon developed good coding standards, they had to step back and identify where maximum returns existed across the entire business. To the annoyance of operations, finance, customer service, and other departments, they determined that commercial (sales and marketing) was the biggest lever they could move.

Increasing sales marginally by 2-4% provided greater returns than improvements elsewhere. As a result, their business improvement team spends 70% of their time supporting the commercial team. Paul acknowledged this is a personal decision for each business, but emphasized the importance of justifying investment by focusing automation efforts where they create the most value, not just where improvements are easiest to implement.

6. Proceed With Extreme Caution on Big Ticket AI Platforms

Kukoon invested in two significant AI platforms: Algolia for web search (£25,000 annually) and Digital Genius for customer service (£35,000 annually). These are big checks for their business, requiring extensive research. They spoke to nearly 20 companies initially, narrowed to detailed inquiries with more than half a dozen, and conducted thorough research including speaking with current customers and seeing live environments.

The critical factor: they required return on investment in less than two years, not just of upfront costs but of time investment. Why?

Because AI technology moves so fast that what’s top of the market one year might be fourth or fifth the next. Paul gave it 50/50 odds they’d still be with either company next year. As ChatGPT creates better API connections, platforms like Digital Genius – which Paul views as having good AI but whose real value is quality integrations – may become obsolete. His assessment was blunt: “There’s a lot of snake oil being sold within the category.” Many companies are just ChatGPT with good connectors, charging tens of thousands for something worth $30 per month. His advice: “Definitely proceed with a lot of caution. There’s not much really, really good out there that ChatGPT can’t do with a good connection.”

7. Automation Benefits Everyone – Even Manufacturing-Heavy Businesses

Paul’s final point was that even if 80% of a business is manufacturing, everyone else is “just moving data and information around the business.” Automating that data movement makes it more accurate, more efficient, more scalable, and provides better customer service. The opportunity for scripts, coding, and automation exists in virtually every business, regardless of industry or operational makeup.

Taking just ten hours to understand what these tools can do represents one of the biggest opportunities for businesses today.