In this section I explore one AI-powered capability and demonstrate how it can be used to create business value. I include concrete examples and lessons learned from actual work with customers, and only recommend approaches that I have used extensively myself and with clients.
One of the most underutilized features in ChatGPT Teams and Enterprise is the ability to create Custom GPTs - specialized versions of ChatGPT trained on your specific documents, processes, and organizational knowledge. While some people know about the public GPT Store where you can find thousands of pre-built assistants, few realize the strategic advantage of building internal GPTs that capture and scale your team's expertise.
I recently worked with a consulting firm facing a common challenge: their senior consultants would conduct intensive face-to-face customer sessions, gathering valuable insights about client needs, pain points, and opportunities, and how they helped them address those. But by the time they returned to the office, much of the nuanced detail had faded. This knwoledge was never captured and was never used to build a reusable knowledge base.
We created a Custom GPT specifically designed to debrief consultants after client consulting sessions. Here's how it worked: immediately after a session, while driving back to the office or their next meeting, the consultant would use ChatGPT's real-time voice feature on their phone to have a conversation with the Custom GPT. The GPT would ask structured questions - "What were the client's primary concerns?" "Which existing solutions did they mention?" "What helped them most during the conversation?", or "What are the key insights from this conversation?"
Through this natural conversation, the GPT captured detailed session notes, identified patterns across multiple client engagements, and summarized insights by industry, challenge type, and solution approach. Over time, this created a living knowledge base of best practices, client concerns, successful responses, and emerging market trends - all captured simply during what was previously dead time.
The power wasn't just in capturing information. It was in making that knowledge immediately accessible and actionable. This newly created knowledge based could now be used in many different ways - supporting less expereinced consultants, creating content for customers based on best practices, and building new product offerings based on this klnowledge and AI-powered capabilities.
Custom GPTs work particularly well for:
Policies and guidelines: Instead of forcing employees to dig through SharePoint for your expense policy or compliance checklist, create a GPT that answers questions conversationally and provides relevant examples based on past scenarios.
Processes and SOPs: Document your standard operating procedures in a GPT that can walk team members through complex workflows step-by-step, answering questions about edge cases and exceptions as they arise.
Templates and frameworks: Rather than distributing static templates, build GPTs that guide users through completing documents interactively, ensuring quality and consistency while teaching best practices.
In the video below, Charlie Gower demonstrates exactly how to build Custom GPTs for your organization - from setting up instructions and uploading knowledge documents to sharing them with specific teams. He walks through a practical example of a "Prompt Writer GPT" that helps employees create better prompts by guiding them through a structured template, then generating downloadable documents.
To see a Custom GPT in action, I've built one based on Amazon's Working Backwards methodology that I've covered in previous issues of this newsletter. The Working Backwards Assistant acts as an expert facilitator, helping you work through customer-centric ideation and draft press releases that start with customer needs rather than solutions.
Whether you're exploring new product ideas or want to apply the Working Backwards framework to your own initiatives, this GPT guides you through the process with the same rigor Amazon applies to product development. You can use it as a template to understand how Custom GPTs structure conversations and deliver value - or simply as a practical tool for your next innovation session.
Identify one area where your team repeatedly explains the same complex information, follows a detailed process, or uses templates that people struggle to complete correctly. That's your first Custom GPT candidate.
Build it, test it with a small group, iterate based on their feedback, then share it more broadly. You're not just creating a tool - you're capturing and scaling your team's expertise in a way that compounds value over time, rather than requiring the same explanations again and again.
The organizations winning with AI aren't just using it for automation. They're using it to capture, preserve, and democratize the knowledge that makes them excellent.