The Setup
My wife heard me typing at the kitchen table. "What are you tippity tapping over there?"
I stood up and explained that I was building a project to extract case studies from my consulting work. A system for turning rich context into consistent, publishable content. We started talking about how colleagues have told her she should document her own process. She runs a travel agency and specializes in using Destination Management Companies to create fully individualized itineraries. The kind of work that's hard to explain but obvious when you see it.
I told her that if she gave me an hour or two to ask questions and record the conversation, I could have her a gated webpage built by tomorrow night. Something she could sell access to.
Great conversation. But then it kind of hung there. I think she was overwhelmed by the idea. I went back to work on my project.
Then she asked: "Did you look at the proposal for shore excursions?"
I couldn't explain what AI could do for her business. But I could show her in 25 minutes.
The Pivot
I pulled up the proposal from her Destination Management Company partner. A package of private shore excursions for a transatlantic cruise. Ten guests, five ports, Portugal and Spain.
She started asking questions. The gist: "Is what I'm putting together reasonable and fair compared to what Virgin offers for excursions?"
A few back and forths later, I said: "Let me drop this into Claude and see if we can't build a quick objective comparison."
The stakes were real. The clients were friends. And another travel agent couple was in the group, along with four of their clients. They would scrutinize every detail. She didn't just need a comparison. She needed proof that her curated package delivered better value than what clients could book directly through the cruise line.
The Revelation
We started cataloguing Virgin's options port by port. Ponta Delgada. Funchal. Tangier. Malaga. Palma de Mallorca. Forty-seven excursions across five ports, with pricing, duration, and descriptions.
A pattern emerged that a price comparison would have buried.
The best excursions were sold out.
Every Sete Cidades tour in the Azores. Gone. The Monte Extravaganza in Madeira. Gone. The private minivan option in Malaga. Gone. The signature experiences that would have made a direct booking worthwhile were no longer available.
This shifted the entire comparison. It wasn't about which option was cheaper. It was about which option was actually bookable. Marissa's guaranteed availability became the deciding factor, not the $224 per person premium.
The Methodology
The browser extension I normally use for web research was disconnected. Rather than troubleshooting, I pivoted. "Do the screenshots of thumbnails work?" I asked her to send me screen captures of the Virgin Voyages excursion listings.
Turns out, thumbnail views gave us exactly what we needed. Price, duration, and availability status in a single glance. Arguably better than scraping individual pages.
What we extracted
| Port | Available | Sold Out |
|---|---|---|
| Ponta Delgada (Azores) | 6 options | 9 options (all Sete Cidades tours) |
| Funchal (Madeira) | 7 options | 4 options (Monte Extravaganza gone) |
| Tangier (Morocco) | 13 options | 2 options |
| Malaga (Spain) | 10 options | 2 options (private minivan gone) |
| Palma de Mallorca | 8 options | 3 options |
The real comparison
Marissa's package came to $749 per person. Virgin's comparable options totaled roughly $525 per person. A $224 difference.
But that $224 premium bought a lot. Full-day private experiences instead of half-day shared tours. Included meals and wine tastings. Guaranteed availability for the signature experiences that were already sold out through Virgin. When you're traveling with friends who are also clients, the difference between "we have the whole van to ourselves" and "we're sharing with 40 strangers from the ship" is worth more than $224.
The gap we found
Marissa's package covered four of the five ports. Tangier wasn't included. Rather than hiding this, we flagged it explicitly and recommended Virgin's best options to fill the gap. Calling out what's missing builds trust. Hiding it destroys it.
The Deliverable
Friends reviewing a travel package don't want a spreadsheet. They want something they can scroll through on their phone while deciding whether to trust the recommendation.
I asked Claude to build an interactive HTML comparison with horizontally scrolling day-by-day cards. Each card showed both options side by side, with winner badges and key differentiators. A summary section at the top with the verdict. A differentiators grid at the bottom.
The summary comparison with verdict banner. Built in 25 minutes.
The deliverable did the selling. When she showed it to her clients, the response was immediate. The comparison was clear, the reasoning was transparent, and the recommendation was backed by evidence they could verify themselves.
The Principles
What transfers to other work:
When the browser extension didn't connect, we didn't troubleshoot. We found another path. Screenshots of thumbnail grids gave us price, duration, and availability in one glance. The goal is the deliverable, not the method.
The "best" option is meaningless if it's sold out. Surface availability status early. Every Sete Cidades tour on Virgin was gone. That made Marissa's guaranteed booking the only way to get the Azores' signature experience.
Wholesale rates in foreign currency don't compare to retail prices in dollars. Make sure you're accounting for conversion, fees, and margins before declaring a winner. The sticker price is never the real price.
Friends reviewing a travel package don't want a spreadsheet. They want something visual and scrollable that respects their time. The HTML artifact with day-by-day cards tells the story better than any comparison table.
Incomplete proposals lose trust. Calling out what's missing, like the Tangier coverage gap, with a recommendation to fill it shows thoroughness. It's better to surface the weakness yourself than have someone else find it.
The Meta-Lesson
An hour earlier, I'd tried to explain what AI could do for her business. Document your process. Build a gated webpage. Sell access to your methodology. She heard me, but the idea hung there. Too abstract. Too much to take in.
Then a simple question about shore excursions became a live demonstration. In 25 minutes, she watched the comparison take shape. She saw the sold-out pattern emerge. She watched the deliverable come together in real time.
That's worth more than any explanation.
The people in your life who are skeptical about AI, or overwhelmed by it, or unsure where to start, are not going to be convinced by your enthusiasm. They're going to be convinced by a moment when you solve their problem faster than they thought possible.
Find their shore excursion question. Then drop it into Claude.
"What are you tippity tapping over there?"
The question that started everything
