Before we ever build anything, we run something close to an audit: what's actually eating your time, what's costing you leads, and what's worth fixing first. You don't need us in the room to start that process. Here's the actual prompt, free to copy and run in ChatGPT or Claude right now.

It won't build anything for you. What it will do is turn a vague feeling of "we're drowning in busywork" into a specific, ranked list you can act on, which is most of the hard part.

The prompt

Copy this whole blockYou are a business process consultant helping a small business owner find their highest-value automation opportunity. Ask me one question at a time, waiting for my answer before the next question, covering: 1) what industry I'm in and roughly how many people work here, 2) the three tasks that eat the most time every week, 3) which of those tasks is most repetitive and rule-based versus which requires real judgment, 4) what happens today when a lead or customer inquiry comes in after hours, 5) how leads or customer data get tracked right now (spreadsheet, CRM, memory, etc.). Once you have my answers, produce a ranked list of the top 3 automation opportunities, in order of highest impact to lowest effort to implement. For each one, explain in plain English: what's happening now that's costing time or money, what a fixed version would look like, and a rough estimate of hours saved per week. Do not recommend anything that requires me to learn new software I don't already have. Be honest if none of my answers point to a strong automation opportunity yet.

Paste this into a fresh chat. Let it interview you: the back-and-forth is what makes the output specific instead of generic.

Why this works better than "how can I automate my business"

Ask an AI assistant that question directly and it'll hand you a generic list: use a CRM, automate emails, try a chatbot. None of it is wrong, and none of it is useful, because it isn't based on anything about your actual business.

The prompt above forces the AI to gather real information first, the same five things worth checking before automating anything, then reason from your specific answers instead of generic best practices. That's the difference between a rough direction and something you could actually act on this week.

What to do with the answer

Two honest outcomes here, and both are useful:

  • It surfaces something concrete and specific. Good. You now know exactly where to start, whether that's doing it yourself, hiring for it, or having it built properly.
  • It confirms nothing is urgent yet. Also good. That's a real answer, not a failure, and it saves you from automating the wrong thing just because you asked an AI to find something.

If the output points to something specific and you want it actually built (not just diagnosed), that's the exact conversation a free consultation with Goldenspire is for. Bring the AI's output with you: it's a solid starting point for a 30-minute walkthrough of whether it's worth building and what that would look like.