Why ChatGPT Changed the Rules
When OpenAI launched ChatGPT in November 2022, few of us imagined the impact it would have on business. Two years later, it is an essential tool in thousands of businesses.
But implementing it properly is another story. I have seen businesses pay for a subscription and barely use it, while others have completely transformed their productivity.
The difference is in the implementation.
Implementation Options
Option 1: ChatGPT Plus ($20/month per user)
Ideal for: Individual use, small teams
Advantages:
- Access to GPT-4
- Plugins and custom GPTs
- Code Interpreter
- Integrated DALL-E
Limitations:
- No enterprise privacy guarantees
- Data may be used for training
- No SSO or user management
Option 2: ChatGPT Team ($25/month per user)
Ideal for: Teams of 3–149 people
Advantages:
- Shared workspace
- Data is NOT used for training
- GPTs shared within the team
- Administrator controls
Option 3: ChatGPT Enterprise (Custom pricing)
Ideal for: Businesses with strict security requirements
Advantages:
- SSO with SAML
- Unlimited GPT-4 use
- Extended context window
- API credits included
- SOC 2 compliance
Option 4: OpenAI API (Pay as you go)
Ideal for: Integration into your own systems
Approximate prices:
- GPT-4 Turbo: $0.01 per 1K input tokens, $0.03 output
- GPT-3.5 Turbo: $0.0005 per 1K tokens
Use Cases by Department
Sales
- Personalised email generation
- Prospect research
- Presentation preparation
- Analysis of common objections
Marketing
- Social media content generation
- Campaign ideas
- Advertising copywriting
- SEO: keyword research
Customer Service
- Response templates
- Summaries of long conversations
- Ticket translation
- Dynamic FAQs
Operations
- Process documentation
- Data analysis with Code Interpreter
- Report generation
- Workflow optimisation
Legal/Compliance
- Contract summaries
- Identification of key clauses
- Version comparison
- (Always with human review)
Prompt Engineering: The Key to Success
80% of the value you get from ChatGPT depends on how you formulate your prompts.
The RACE Structure
Role: Define who ChatGPT is in this context Action: What you want it to do Context: Relevant information Example: Show the expected format
Role: You are a B2B email marketing expert with 10 years of experience.
Action: Write a follow-up email for a prospect who did not respond to my first proposal 1 week ago.
Context:
- Business: Automotive parts importer
- Contact: Purchasing Manager
- Proposal: CRM implementation ($5,000)
- The first meeting was very positive
Example of the tone I want:
"Hello [Name], I hope you have had an excellent week..."
Custom Instructions
Configure instructions that apply to all your conversations:
"About me":
I work at a technology consulting business in Ecuador.
My role is CTO.
Our clients are SMEs with 10–100 employees.
I prefer direct, practical answers.
"Response preferences":
- Respond in Latin American Spanish
- Use examples of real businesses where possible
- Include pros and cons in recommendations
- Format with bullets and headings for easy reading
Custom GPTs for Your Business
GPTs are customised versions of ChatGPT trained with your instructions and knowledge.
Examples we have created for clients:
Onboarding GPT:
- Knows all the company's processes
- Answers frequently asked questions from new employees
- Provides step-by-step guidance for common tasks
Sales GPT:
- Has the company's pitch
- Knows common objections and responses
- Generates tailored proposals
Technical Support GPT:
- Product documentation uploaded
- Troubleshooting guides
- Escalates to humans when needed
Security Considerations
Data you should NEVER share with ChatGPT (free version):
- Customers' personal information (names, ID numbers, emails)
- Sensitive financial data
- Passwords or credentials
- Confidential intellectual property
Best practices:
- Anonymise data before sharing it
- Use ChatGPT Team/Enterprise for sensitive data
- Train your team on what is and is not appropriate to share
- Always review responses before using them externally
Suggested Implementation Plan
Weeks 1–2: Pilot
- Select 3–5 early adopters
- Configure ChatGPT Team
- Define 3 initial use cases
Weeks 3–4: Training
- Prompt engineering workshop
- Document the best prompts
- Create the first GPTs
Month 2: Expansion
- Roll out to more departments
- Collect feedback
- Optimise GPTs
Month 3+: Optimisation
- Measure impact
- Integrate with other tools
- Explore the API for automations
Need help implementing ChatGPT in your business? Contact us for a tailored plan.
