From 80 Pages to 5 Minutes

The workbook worked. Nobody finished it.

80 pages. PDF. Fill it out yourself. The results were transformative — when people actually got to the end. But the completion rate was brutal. Too long, too complex, too much friction.

The solution wasn't obvious. It took three attempts.

Attempt 1: The Workbook (February 2025)

80+ pages. PDF. Self-guided. Results: fantastic. Time required: days. Completion rate: dismal.

Verdict: the content was right. The format was killing it.

Attempt 2: Custom GPT (April 2025)

I packed the entire workbook into a Custom GPT. "Your Reflection" — a radical mirror that asks the uncomfortable questions.

"It probes. Asks. Confronts. Until you get honest."

Better. Interactive instead of static. Minutes instead of days. But GPT has limits. No backend. No data. No control over the prompt. And only people with ChatGPT Plus could use it — everyone else was locked out.

Verdict: the format was right (conversation). The platform wasn't.

Attempt 3: My Own Chatbot (March 2026)

February 12, 2026: bootstrapped the Booster repo. March 8: first chatbot commit. March 10: live at diagnose.alexreisch.com.

Stack: Next.js, Claude API, Supabase, Stripe. All custom. All under my control.

226 commits. 184 hours. 40 sessions with my AI partner.

What would have taken a team of five — developer, product manager, copywriter, marketing, community manager — I built alone. With AI. In 32 days.

What the Chatbot Does Differently

It asks the same questions as the workbook. But it adapts them. In real time. Based on your answers. No static PDF that treats everyone the same — a conversation that shapes itself around you.

5 minutes. Not 5 days.

The mirror principle is the same: show people their blind spot. In their language. Without the fluff. Just faster. And accessible to anyone.

The First Numbers

After 2 weeks live:

• 45 sessions
• 52% made it to the diagnosis
• Average duration: 5 minutes
• 60% had the same blind spot: positioning

60%. Six out of ten coaches, consultants, and experts who used the chatbot had the same core problem. They weren't speaking their audience's language. Exactly like Henning said on that call: "I have a good offer, but I'm talking past my target audience."

The workbook uncovered that in days. The GPT in minutes. The chatbot in 5 minutes — and it collects data that shows me where the next lever is.

The Evolution

Workbook (days) → Custom GPT (minutes, but limited) → Own chatbot (5 minutes, deep, personal).

Three attempts. One core. Getting better each time.

And the best part: I'm not done. Version 2 of the prompt is already live. A/B tests are running. The data shows me every day where I can improve.

80 pages to 5 minutes. That took a year. But sometimes you need a year to understand that the result was right all along — just not the packaging.

— Alex, Paphos, March 2026

---

226 commits. 3 iterations. 1 core.

From workbook to GPT to my own system. The format changed three times. The question stayed the same: where's your blind spot? Today an AI answers it in 5 minutes.