I just used Claude Code + Skill Creator to build a small opportunity research system designed to find AI automations that could realistically be sold to Italian SMEs.

Instead of doing random brainstorming, I set Claude up to think more like a vertical automation consultant.

Here’s the process I followed, step by step:

1. I defined the goal clearly.I wanted to identify real automation opportunities for Italian businesses, not vague ideas or generic startup concepts. The goal was to find repetitive, expensive, operational problems that could justify an offer in the €3k–€5k range.

2. I created a set of specialized skills.Using Skill Creator, I built multiple skills, each with a specific role:

one to identify business pain points
one to turn those pain points into automation opportunities
one to design the technical workflow
one to build the commercial offer
one to identify the types of companies that would buy it
one to define a fast MVP

So instead of using Claude in a generic way, I gave it a more structured operating system.

3. I upgraded the skills to make the output more sellable.I forced three key elements into the responses:

ROI calculation
pricing model
vertical business focus

That matters because a lot of AI output sounds smart but is not directly usable in sales. I wanted responses that already contained the economic logic a client would care about.

4. I tested the skills with internal benchmarks.I compared “with skill” vs “without skill” outputs, and the difference was clear: with the skills enabled, Claude produced more complete workflows, added pricing, included ROI, recommended concrete tools, and followed a more structured logic.

5. I ran a final prompt focused on the Italian market.I asked Claude to find high-value automation opportunities in Italy across industries such as:

dental clinics
real estate agencies
accounting firms
gyms
marketing agencies
restaurants / hospitality

6. I got a final summary table of ranked opportunities.The result was a table including:

industry
solution
setup fee
monthly retainer
monthly client value
payback period

Some of the strongest opportunities that came out were:

appointment reactivation and reminders for dental clinics
AI lead qualification pipelines for real estate agencies
document collection automation for accounting firms
member churn prevention for gyms
automated client reporting for marketing agencies
review management for restaurants

7. Final insight.The most interesting takeaway is that dental clinics and real estate agencies seem to offer the fastest and clearest ROI. The pain is easy to understand, the business case is obvious, and the payback period is relatively short. That makes them strong verticals for first offers and first case studies.

To me, this is a strong proof that with the right skills, Claude Code is not just a coding assistant. It can become a real engine for discovering, structuring, and validating automation offers that are actually sellable.

If anyone here is building something similar, I’d love to compare workflows, prompts, or vertical strategies.

1 comment