7 Hours vs. 4 Months
Why Your Next IT Project Costs $35K Too Much
Tuesday morning it was a stray thought. Thursday morning it was in the hands of a doctor who said: “This actually holds up.” Seven hours. The same amount as a working day full of meetings, coffee and answering pointless emails. Or it’s the time it takes to build a complete, evidence-based health app from scratch. But only when you stop having meetings and start building.
I’m done talking about “AI potential.” We’ve sat through enough PowerPoints about disruption. I wanted to prove that the future is here now. That as a modern leader, you can go from idea to market in the time it used to take to write the first sketch of a project specification.
This isn’t about health apps. This is about the barrier to innovation collapsing. For SMBs, your biggest enemy isn’t the competition—it’s the time you spend wondering if you dare.
The app in hand
The Project: An Antidote to “Adult Gummy Bears”
Why health? Because I’m allergic to bullshit. We’re currently paying a “wellness tax” that’s tangible. The average Dane drops DKK 1,500 per month (200 USD) — that’s DKK 15,168 annually (2,300 USD) —on protein powder, expensive fitness club memberships, and vitamin gummies that are essentially Haribo for adults.
I wanted to build Vitality40+. An app based on hard science from WHO, Mayo Clinic and Harvard. An app that won’t sell you anything, but reminds you that your body already has the world’s best detox machines built in: your liver and kidneys. They don’t need juice cleanses; they need the right working conditions.
The Engine Room: 7 Hours, 3 Tools, 1 Mission
Development happened over three days (because I actually take weekends off). Here’s the technical chronology:
Phase 1: Research (3 hours)
I didn’t start by guessing. I threw a comprehensive BBC article on biology into Gemini Deep Research. In 20 minutes, I had a validated catalog of the 40 most effective “Exercise Snacks.” The science is clear: just 10 extra minutes of movement per day can statistically prevent tens of thousands of deaths annually for people over 40. The raw version needed cleaning for relevance and authority. The remainder became part of the spec to guide generation of exercise snacks.
A small part of the summarized output and a ton of research links.
Phase 2: Execution (3 hours)
I used Gemini 3 Flash to write my PRD. Then it went into Google Stitch for four design iterations and pre-testing of user attention and, finally, Google AI Studio for code. The result? A lightning-fast Progressive Web App (PWA) with Privacy-by-Design (everything stored locally on your phone, zero data harvesting) and full internationalization.
Google Stitch for useful and pretty designs showing version 4 of the final design
Phase 3: The “Brutal” Feedback (1 hour)
Four test users—including a doctor and an AI developer—gave feedback. In 60 minutes, I prioritised the input, updated the spec, added the functionality and re-tested:
Silent Bias: We implemented gender and age logic, so the app automatically weights bone-loading exercises for women over 50 (bone health/menopause).
Intensity Levels: From “desk-friendly” to Level 3, where your heart rate actually climbs.
Mental Resilience: We built in Box Breathing and knowledge about sleep as the brain’s “waste management system.”
The stuff that normally takes two weeks? Done in an hour. Not because I’m brilliant, but because the AI was open while the feedback was still as warm as the coffee.
The Business Case: Decision Paralysis is Your Biggest Expense
How many times have you killed an idea because the IT vendor said: “$35K and four months”?
This isn’t about health apps. This is about the barrier to innovation collapsing. For SMBs, your biggest enemy isn’t the competition—it’s the time you spend wondering if you dare.
With this approach, you can:
Test a hypothesis in two workdays: Stop guessing what customers want. Build it and watch them use it.
Prototype for $0: Remove the financial risk in the early phase. Total cost of Google Cloud usage is currently less than a cent.
B+ in market beats A+ on the drawing board: My app isn’t perfect, but it’s live, doctor-validated, and working.
Three Truths You Can Use Tomorrow
1. AI executes. You think.
I still had to choose which feedback mattered most. AI built what I pointed at. It’s not magic—it’s a power tool.
2. Stop looking for one tool.
I jumped between Gemini, Stitch, and AI Studio like a carpenter switching between hammer and saw. Multi-tool workflows are the new normal.
3. Speed kills paralysis.
The app I shipped on Thursday is better than the perfect app that’s still in planning on Friday.
Start Now—Or Let’s Do It Together
The world doesn’t need more plans. It needs execution.
If you have a process that never got digitized, or a customer journey you know could be better, stop putting it on next quarter’s agenda.
We can meet, open our laptops, and build your prototype while we sit there. We don’t leave the meeting with a plan. We leave with something that works.
Coffee? 7 hours? Let’s see what we can create.
Contact me at: frimer-rasmussen.dk
Try Vitality40+
P.S. The app took 7 hours to build. How long did your last IT project take—and did you get what you asked for?