Prompt engineering might be the new hawtness but it is not a business.

Prompt Engineering is Necessary but Not Sufficient.

Here's the truth: prompt engineering is a foundational skill in the AI world. It’s how to talk to AI. It's crucial, it's powerful, but it's not a business by itself.

Hell, I have a trademarked prompt engineering framework (RISEN™) framework and even so I’m telling you prompt engineering is just a small part of the puzzle.

Can you sell prompts? Sure, kind of. Lots of people do. But let's be real – a raw prompt collection is a low-quality asset. They are mostly crap. You've probably downloaded a few yourself (and promptly ignored them, pun intended).

The real value, the place where the rubber hits the road, is in workflows. Specifically, stringing together prompts to help a business DO SOMETHING.

Let’s get started:

The Two-Stage Approach: Identify and Implement

Your job as an AI Workflow Consultant breaks down into two crucial stages:

1. Identify: This is where your business skills shine. Bring your industry knowledge to the table. Find processes that companies want to optimize, areas where they're looking to reduce costs or save time. The key is to systematize what matters, not to automate for the sake of automation.

2. Implement: This is where your AI skills come into play. Once you've identified the opportunity, you'll use your AI knowledge to create solutions.

Remember that our skills as AI entrepreneurs come from marrying business and AI. This is true here and beyond.

Let's break these down further.

Step 1: Identify - The Business Skill

The foundation of great consulting is asking the right questions. To help you get started, here's a prompt to generate a framework for client interviews:

You are an expert business process consultant. Create a framework of questions to ask a client to understand their business needs and identify processes ripe for AI optimisation. The questions should help uncover:

1. Current inefficiencies and pain points
2. Key business objectives
3. Existing workflows and processes
4. Input and output requirements for each process
5. Resource allocation (time, money, personnel)
6. Decision-making bottlenecks
7. Data management challenges

Provide at least 3 questions for each category. Present the questions in a logical order that would flow naturally in a client interview.

Use the output from this prompt as a starting point for your client interviews. Remember, the goal is to understand the inputs, processes, and outputs of their current workflows.

Pay close to attention to suggestions like this:

Because you already know your industry (you focused on an industry you already know right? Right?) you can adapt these questions to your industry. You probably already have a fair idea of the problems that can be solved but always (always) ask your customers. We tend to get this part wrong!

When you sit with clients you’ll want to listen and be sketching out what these processes might look like. Then we’ll use this in the next step.

Step 2: Implement - The AI Skill

Once you've identified the processes to optimise, it's time to put your AI skills to work. But here's a secret: you don't need to build everything from scratch.

For now, you can do an awful lot with off-the-shelf automation tools like Make (formerly Integromat) and Zapier. The basic idea is to chain prompts and software together to create efficient workflows.

Here’s what a workflow “looks like” in Make for example:

Your job is to identify problems (as above) and translate that into automation workflows like the one above.

The very basic idea here is to take an input, run it through a process (or processes) to create a output. Your success comes from simplifying this flow as much as possible so that an AI can run the workflow.

To help translate client needs into actionable processes, use this prompt:

You are an AI workflow designer. Based on the following customer problem, inputs, and desired outputs, suggest a high-level workflow that uses AI and automation tools to solve the problem:

Customer Problem: [Insert problem here]
Available Inputs: [List available data or resources]
Desired Outputs: [Describe the ideal end result]

Your workflow should:
1. Identify key steps in the process
2. Suggest specific AI tools or models for each step (e.g., GPT for text generation, computer vision for image analysis)
3. Highlight any potential challenges or considerations

Present your response in a clear, step-by-step format that a non-technical business owner could understand.

This prompt will take the customer problem and map out the workflow for you. It’ll do the heavy lifting. We’re using AI to suggest AI processes. Cause we’re smart! 😋

Pro Tip: Zapier has a built-in tool to suggest workflows. Don't reinvent the wheel – use it as a starting point and then customize based on your client's specific needs.

Let’s say that we have a specific input/process/output step like this:

We’d ask our AI to shorten it (less than 400 characters) which gives us:

Automate customer communication by integrating Shopify with Klaviyo for real-time order status updates (confirmation, shipment, delivery). Use OpenAI's API within Klaviyo to generate personalised messages based on order history and customer preferences.

I can now plug that into Zapier’s AI suggestion tool and it builds the workflow for me:

Work from here to refine and then test, test and test more. Still work to be done but we’ve dramatically reduced our workload.

Building Your Consulting Brand

Now that you've got the skills, it's time to showcase them. Your content should demonstrate your ability to solve real business problems with AI workflows.

Try this:

1. Create case studies of process optimisations (hypothetical or, even better, real)

2. Write articles comparing different AI tools and automation platforms for specific business tasks

3. Start a YouTube channel demonstrating AI workflow optimisations

4. Host webinars on "AI-Powered Efficiency for [industry]: Streamlining Your Business Processes"

Remember, you're not just sharing information now – you're demonstrating expertise and problem-solving skills. People will not just go off and DIY. They’ll instead hire you.

How to get started with all this?

1. Reach out to a local small business or non-profit

2. Offer to analyse and optimise one of their processes for free in exchange for you writing and publishing about the process

3. Document the entire process, from analysis to implementation

4. Create a case study showcasing the results

5. Collect a testimonial

6. Publish it all!!!

This isn't just practice – it's the start of your portfolio. Proven results are key.

What's Next?

In the next part, we'll dive into Stage 3: AI Builder. We'll explore how to create standalone AI tools that solve specific business problems. This is where you'll start to move beyond off-the-shelf solutions and into custom development.

Remember, every level builds on the last. Your skills in identifying business needs and creating AI workflows will be crucial as you move into building your own AI solutions.