Few Lessons from a failed startup

By Slobodan Ninkov

Context: I tried to build a startup based on “simulated field usage and testing”. The need is real: in product development with field testing, simulation can be far cheaper than running systems in the field. Case Study: Simulation-First Validation

This experience strongly influenced how I think about analytics and data work today: clarity beats cleverness, incentives shape behavior more than intent, and systems fail at their boundaries, not their centers.

What I learned:

Paperwork is real

Starting, running, and closing a company means lawyers, auditors, and filings.

They’re friendly—but they get paid whether you sink or swim. Budget for their services. As a limited company, expect at least annual director responsibilities and audits.

Be clear: product or service

Product = you sell a thing. Service = you sell time/knowledge. I tried “product with customization.” It muddied the message and confused customers.

The EU favors products

Services don’t scale easily; policy and incentives often lean toward product companies.

GDPR is not optional

If your site has logins, you’re under GDPR. Competitors may report you. Compliance and cybersecurity are real costs and responsibilities—almost a separate business function.

Anything near defense has its own gravity

Different rules, long procurement cycles, and relationship-driven. The system favors big, slow incumbents. Connections matter.

Distribution beats almost everything

If you crack marketing and sales, you can rent development and still build a product.

TAM isn’t what you think

Your “imagined” market and your “accessible” market are different. Validate the reachable slice.

Technical excellence ≠ traction

Without sales and marketing, craftsmanship doesn’t move the needle.

!!! Important takeaways:

Nail the product messaging. Clear > clever.

Make the value obvious to people who aren’t already problem-aware. “This is how we do things” inertia is your real competitor.

If you’re working on something similar (simulation, field testing, regulated markets), I’d be happy to compare notes.