Prompt Playbook: AI Automation Business PART 2

Hey Prompt Entrepreneur,

Some people struggle to see the treasure right in front of their eyes.

Here’s a powerful question.

"What's the most annoying part of your job?"

When asked this most people (understandably) vent:

"God, where do I start? I spend 8 hours a week copying client meeting notes into our CRM. Then there's the compliance reports where I have to gather data from five different systems. And don't get me started on creating those custom portfolio summaries for each client..."

THESE are the sort of problems we should be fixing with AI automation.

People think automation means building something revolutionary when the most profitable solutions are hiding in their daily irritations.

Because guess what - if you have these problems then hundreds, nay, thousands of other people do.

Your industry expertise isn't just helpful for building automations—it's your secret weapon. This is what gives you the edge. While others struggle to understand business problems from the outside, you already know exactly where the pain points are and what they cost. And this is invaluable.

Let’s get started:

The Gold in Your Daily Frustrations

Every time you think "I hate doing this" or "there's got to be a better way," you've identified a potential automation.

But we become blind to our own problems. The tasks that drain our time become so routine that we stop noticing them. That's like a fish asking "what water?" We swim in these inefficiencies daily without recognising them as business opportunities.

The longer you put up with these annoyances the less you notice them. But they are still annoying!

The examples I gave above might be tens or hundreds of thousands worth of annual automation opportunities. We just might not recognise them as such because they aren’t "AI problems"—they are just our daily problems.

This is your unfair advantage over generic automation builders: you don't need to research or guess at problems. You live them. You know exactly how much time they waste, what they cost, and who makes the buying decisions.

BUT … we need to extract them from you. Because you may be inured to them by now.

The Boring Problem Framework

So what are we looking for?

When hunting for automation opportunities in your industry, focus on tasks that are:

Repetitive without being creative. You want processes you do the same way every time, not tasks requiring judgment or creativity. Copying data between systems? Perfect. Negotiating a custom deal with a client? Maybe not yet.

Time-consuming but not complex. If it takes hours but follows clear rules, it's automation gold. If it requires expertise that changes based on context, save it for later. We like simple!

Costly in terms of human time. Look for tasks where someone's valuable expertise is being wasted on manual work. An accountant doing data entry is expensive data entry. Their time should be used elsewhere on more productive tasks and the mundane automated.

Clear input, process, output. The best automations have obvious triggers (ie. form submission), defined processes (ie. validate and format), and specific outcomes (ie. create record and send email).

Here's a prompt that will use the interview method to extract this information from you:

You are a business efficiency consultant specialising in finding high-value automation opportunities. Interview me to extract specific problems from my professional experience, focusing on the most boring, repetitive tasks that waste expensive time.

Focus on identifying problems that are:
- Repetitive without being creative
- Time-consuming but not complex
- Costly in terms of human expertise being wasted on manual work
- Have clear input, process, and output steps

Ask questions one at a time, focusing on:
- Tasks I do repeatedly that take 30+ minutes each time
- Information I move between different systems or formats
- Processes where I do the same thing every time without variation
- Work that requires no creative input but wastes my expensive time
- Situations where small errors cost time but aren't critical

For each problem identified, help me estimate:
- Time spent per week
- Number of people affected
- Clear trigger points and predictable outcomes
- Whether this follows the same pattern each time

Begin by asking about my industry and role, then dig into those tedious processes that make me groan every time they appear on my to-do list.

Take 30 minutes to run through that interview prompt. Don't edit yourself—just capture every frustrating process you can think of. The purpose is to create a long list at this point.

The Red Flags to Avoid

Now we want to take our long list and slash and burn it. We’ll do a first rough pass removing any of these red flags:

The "nice to have" trap. If people tolerate the current process without major complaint, they probably won't pay much to automate it. Look for problems that genuinely cost significant time or money.

The "already solved" market. Avoid problems with 10 existing solutions unless you have a dramatically different approach. Competition isn't necessarily bad, but saturation is. For our first automation at least let’s find something relatively unique.

The "every instance is unique" challenge. Tasks where each case requires different handling are tough to automate. Focus on standardised processes with predictable variations.

The "requires human judgment" barrier. Automations excel at following rules, not making decisions. Save the complex judgment calls for later projects. We can add steps into our automation that ask for human intervention but again it’s additional complexity we don’t need right now.

Filter Step

Once you’ve removed the easy duds we should have a much shorter list. Now we want to more carefully analyse through six critical lenses.

Make a sheet a grade each problem on these areas:

Time Impact: Don't underestimate this. That "quick" task you do three times a week? If it takes an hour each time, you're looking at 150+ hours annually. Scale that across a department and suddenly you're talking about thousands in labor costs.

Financial Impact: Think beyond just time. What about the downstream effects? Delayed responses to clients, missed deadlines, errors from manual data entry—these cost real money. A single data entry mistake can easily cost more than an entire automation solution.

Market Size: This determines if you're building a one-off solution or a scalable business. Is this a problem specific to accounting firms with 10-50 employees? Or do all service businesses struggle with this? The more universal the problem, the bigger your potential market.

Technical Feasibility: Ask yourself: can this be broken down into trigger, simple AI processing, and output? If you're thinking about multiple decision points, exceptions, and custom rules—that's probably too complex for your first automation. Look for the straightforward paths first.

Competitive Landscape: Existing solutions aren't necessarily bad—they validate demand. The key question is: what's missing? Are current solutions too complex? Too expensive? Don't handle your specific industry's needs? These gaps are your opportunities.

Sellability: The easier it is to demonstrate value, the easier it is to sell. "This saves 8 hours weekly" is more sellable than "This optimises your workflow." Concrete, measurable benefits close deals.

Use these six factors to rank your problems. The winners? High time and financial impact, clear technical feasibility, and obvious sellability. Market size determines if you start with custom solutions or aim for a productised service.

FYI: I go into this in more depth in the AI Automation Accelerator (as well as providing analyse prompts) but this will get you started along the right track.

What's Next?

Next up we'll take these identified problems and transform them into specific automation blueprints.

I'll show you how to use AI to design clear workflows without overcomplicating things—the difference between products that actually get built and just another idea stuck on a to-do list.

Keep Prompting,

Kyle

If you feel this — learning how to use AI in entrepreneurship and work — is not for you → Unsubscribe here.