How to Train AI to Work Smarter: A Proven Business Process

Struggling to make AI work effectively for your business? Want to know how to create AI assistants that truly understand your unique needs?

In this article, you'll discover a proven process for building a dream team of custom GPTs to elevate your marketing efforts and streamline your operations.

Understanding the AI Implementation Journey

Initially focusing on consulting for the non-profit sector during the early days of social media, Kate vanderVoort quickly recognized how technology could help organizations better connect with their audiences, and she has maintained a deep passion for understanding the intersection between technology and humanity throughout her career.

Her experience with artificial intelligence (AI) predates the ChatGPT phenomenon, having worked with early platforms like Jasper.ai and Copy.ai. However, witnessing ChatGPT's capabilities sparked a deeper exploration of how AI could enhance every aspect of business operations. This background gives her unique insight into the current state of AI adoption and implementation.

Despite the rapid advancement of AI technology, vanderVoort emphasizes that we're still in the early stages of the AI transformation. “For people like us, we're in this all day, every day, and sometimes I forget that there are still people that haven't yet even heard about ChatGPT or had their ChatGPT moment,” she explains. She notes that marketers, in particular, are at the forefront of this technological wave.

The speed of change in AI technology cannot be overstated. Kate consistently reminds her workshop and webinar participants that the AI landscape they're looking at today will likely be completely different in 12 months – sometimes even in as little as 12 days. This rapid evolution makes it crucial for businesses to develop flexible, adaptable approaches to AI implementation.

Creating CustomGPT Dream Teams of AI-Powered Virtual Assistants

Many businesses are underutilizing AI due to fundamental misconceptions about its implementation. According to vanderVoort, the most common mistake is using overly simplistic expectations. Many users input basic AI prompts, receive generic outputs, and then dismiss AI as inadequate for their needs. This surface-level interaction fails to tap into AI's potential for business enhancement.

Another critical misconception is viewing AI solely as a replacement for human work rather than an enhancement tool that should support employees with repetitive tasks. “One of the concepts we talk a lot about is how do we bring the best of who we are and bring our expertise and our knowledge and then look at how AI can enhance, can help us to do more, serve more people, have a greater impact, rather than how do I bring my task and just outsource it to AI,” vanderVoort explains. This perspective shift is essential for successfully embracing AI.

The custom GPT dream team concept represents a transformative approach and brings two key benefits to AI implementation in business. 

• Rather than relying on generic AI interactions, these specialized assistants serve as dedicated team members, each trained to understand and enhance specific aspects of your business operations. 
• These AI assistants offer unprecedented availability, operating 24 hours a day across all time zones. This constant availability means businesses can maintain consistent operations and support regardless of time zone constraints or working hours.

Remember: Everybody's dream team looks different because we've got different strengths, do different tasks, and target different audiences.

Step 1: Conduct a Comprehensive AI Audit

The journey to enhancing productivity by building an effective custom GPT dream team begins with thoroughly understanding your current business operations. Kate's approach to this crucial first step varies depending on organizational size and complexity, but the fundamental principles remain consistent.

For larger organizations, her team conducts extensive AI business audits that examine every task, role, and department across the business. This comprehensive analysis provides a benchmark for the aggregated value and time investment in various tasks.

Small businesses and solopreneurs can adopt a more streamlined approach to this audit process by documenting activities for at least a month, as many tasks occur on monthly rather than weekly cycles. This documentation doesn't require complex tools – a simple Excel spreadsheet, time-tracking app, or even conversations with ChatGPT can effectively capture the necessary information.

Through this process, organizations often uncover surprising insights about resource allocation with regard to time spent on mundane tasks. The key is to create a detailed record of tasks that are repeated daily, weekly, and monthly in your business. This documentation serves as a baseline for measuring future improvements and helps identify prime candidates for AI enhancement. Marketers particularly appreciate having this benchmark to measure their return on investment in AI implementation.

AI Implementation Case Study

A non-profit organization Kate works with operates a residential facility where participants complete handwritten feedback forms weekly. The traditional process involved an individual employee manually performing data entry by entering the details from 40-50 handwritten forms each week, consuming approximately six hours.

Through careful analysis and AI implementation, vanderVoort's team transformed this cumbersome process to boost productivity. 

The new workflow begins with scanning the handwritten forms into a PDF file, which takes just minutes. The AI then transcribes the handwritten text, automatically generates an Excel spreadsheet, and performs sophisticated analysis, including sentiment evaluation and testimonial extraction. The system even automatically drafts email responses for management to review before sending them to feedback providers.

The result dramatically reduced processing time from six hours to just 20 minutes per week. More importantly, the quality and depth of analysis improved significantly, providing the organization with better insights into participant feedback while freeing up valuable staff time for other important tasks.

Step 2: Create 10 Foundation Documents to Train AI for Custom GPTs

After your audit, use AI to develop a set of comprehensive foundation documents that will serve as the knowledge base for your AI assistants. 

Kate has developed a systematic 3 tier approach to creating these documents, structured in a way that builds upon and reinforces each preceding document through iterative stacking. This creates a deeply interconnected knowledge base that ensures all aspects of your business strategy remain aligned and consistent.

Pro Tip: Save these documents as text files or Word documents rather than PDFs, as they're easier to update and tend to work better with AI systems.

Tier 1 Documents

This first document captures everything about you as a professional – your background, experience, educational history, personality profiles, strengths, and weaknesses. For American audiences, vanderVoort notes that this might include more emphasis on academic credentials than might be typical in other markets. The document should also incorporate personality assessments you've completed, such as Myers-Briggs, DISC, or StrengthsFinders profiles.

The document helps AI understand not just your professional capabilities but also your working preferences and limitations. When Kate first implemented this approach, the improvement in AI outputs was immediately noticeable. For instance, vanderVoort openly acknowledges that people management and running large teams aren't her preferred approaches. This self-awareness influences how AI suggests business structure and operational strategies to her.

The second document focuses on your business's structure, industry information, available resources, and operational processes. It builds upon the personal profile, ensuring that AI recommendations align with your individual working style and your business reality.

The third document is a comprehensive market analysis that includes market trends, a SWOT analysis, competitor research, and identification of market opportunities. The AI can then analyze this information in the context of your personal and business profiles to identify unique opportunities that align with your strengths and preferences.

You are a McKinsey-level consultant. Please recommend specific models and frameworks I can use to develop marketing strategies.

Tier 2 Documents

The second section of the foundation documents focuses on your product and service portfolio and target audience personas.

Tier 3 Documents

The final section of foundation documents includes your business plan, financial projections, customer journey mapping, and content strategy. 

This comprehensive understanding becomes particularly valuable for content strategy. If you've identified in your professional profile that you won't create certain types of content—vanderVoort uses the example of “dancing TikTok videos”—then AI won't suggest content strategies that conflict with your established preferences and capabilities. This preference carries through all subsequent planning and recommendations.

Step 3: Decide Where to Implement Custom GPT Assistants

With your foundation documents prepared, the next step is deciding which custom GPTs to create. Upload your initial AI audit results and your business and products/services documents to ChatGPT. You can also use AI tools like Microsoft Copilot, Claude, or Google Gemini.

Then, ask the AI to give you data-driven insights and recommend where GPT implementation would be most valuable for your business based on your regular tasks and business needs. This ensures you're creating AI assistants that address your most significant opportunities for enhancement.

Please review this task tracking sheet and give me data-driven insights and recommendations for implementing custom GPTs to maximize efficiency and effectiveness in my business.

When you have the results, you can create your first custom GPT.

Pro Tip: Building your custom GPT dream team takes time. Kate advocates for an iterative approach, starting with one or two core GPTs and gradually expanding based on results and needs. This allows you to refine your approach and ensure each AI assistant truly adds value to your business operations.

Step 4: Create Your CustomGPT Dream Team

Creating custom GPTs is simpler than many people believe. The process begins in ChatGPT's Explore GPTs section, where you'll find options to either Create or Configure your custom GPT.

For beginners, vanderVoort recommends starting with the Create tab, which allows for a more conversational setup process. However, once you're comfortable with the foundation documents approach, she suggests moving directly to the Configure tab for more precise control.

The configuration process involves several key elements. First, you'll upload an image representing your custom GPT and provide a clear title and description.

Next comes the instructions section, where vanderVoort recommends using the SNAP formula (Set the role, Name the task, Add context, define your Preferred format). The instructions should clearly define the GPT's role, purpose, and preferred communication style while remaining concise and focused; keep your instructions simple and avoid contradictions.

You can even ask the AI to help you create instructions for your custom GPT.

Our team of 10 spends 60 hours a week responding to and writing emails. How can you help us do that better? How can we save time on our emails?

The knowledge files section is where you'll upload your foundation documents.

Create an AI Advisory Board GPT

Begin by talking with ChatGPT about your main business functions. For instance, if you're a consultant, course creator, or content creator, you'd ask ChatGPT to identify the world's leading experts in each of these areas.

I am a consultant who specializes in helping businesses implement AI effectively through hands-on coaching and consulting. I need to create a virtual advisory board to help guide me in making decisions about business strategy, course creation, email marketing, and social media marketing. Please identify the world's leading experts in these fields.

The AI will provide a comprehensive list of experts – for course creation, it might suggest names like Amy Porterfield and Pat Flynn; for social media, perhaps Michael Stelzner; for funnels, Ryan Levesque; and for human-centered AI, someone like Fei-Fei Li from Stanford.

You then create a custom GPT that embodies the collective expertise of these thought leaders, informed by your foundation documents.

You can run ideas for business decisions and content creation past your virtual board and receive feedback from different expert perspectives. The social media expert might suggest improvements to your content strategy, while the funnel expert might recommend specific lead-generation approaches tailored to your business model.

Create a Customer Focus Group GPT

Rather than spending weeks conducting traditional focus groups and surveys, Kate uses AI to simulate focus group discussions. 

The AI creates detailed conversations between fictional customers that reflect your target demographic, discussing their challenges, needs, and reactions to your offerings. This approach saves significant time and provides rich insights into customer perspectives.

Create a Board Member GPT

Another of vanderVoort's innovative applications of this technique was creating an AI board member avatar for non-profit organizations. This specialized AI can review board papers and proposals before sending them to the actual board, providing valuable feedback and insights.

Saving Effective Chats as Custom GPTs

One of vanderVoort's most valuable tips for creating effective custom GPTs comes from leveraging successful ChatGPT conversations. When you have a particularly productive interaction with ChatGPT that achieves exactly what you're looking for, you can ask it to transform that conversation into custom GPT instructions. 

I want to capture everything we've done here so that I can use it again later. Please write me a set of custom instructions for a custom GPT based on this chat.

This approach allows you to replicate successful interactions in a more permanent, specialized format.

Implementation Tips

Document and Prompt Management: For organizations implementing this approach across teams, vanderVoort suggests maintaining a collection of proven prompts that team members can access and customize. This creates consistency in how different team members interact with the AI assistants while allowing for personalization based on specific needs.

Troubleshooting AI Resistance: Sometimes, when attempting to analyze LinkedIn profiles or perform certain tasks, ChatGPT might initially claim it can't help. In these situations, vanderVoort suggests simply reminding ChatGPT that it has helped with similar tasks before. This gentle persistence often results in successful completion of the requested task.

Other Notes From This Episode

• Connect with Michael Stelzner @Stelzner on Facebook and @Mike_Stelzner on X.
• Watch this interview and other exclusive content from Social Media Examiner on YouTube.

Where to subscribe: Apple Podcasts | Spotify | YouTube Music | YouTube | Amazon Music | RSS

✋🏽 If you enjoyed this episode of the AI Explored podcast, please head over to Apple Podcasts, leave a rating, write a review, and subscribe.

---

Stay Up-to-Date: Get New Marketing Articles Delivered to You!

Don't miss out on upcoming social media marketing insights and strategies! Sign up to receive notifications when we publish new articles on Social Media Examiner. Our expertly crafted content will help you stay ahead of the curve and drive results for your business. Click the link below to sign up now and receive our annual report!