The most expensive mistake in this startup was not the code. It was investing roughly RUB 6 million before validating the core unit economics with a real MVP. The most valuable decision was to close before belief turned into denial.
In 2016, my partners and I launched an online-to-offline beauty service: stylists, hairdressers and nail specialists on demand. The Russian market appeared open, while traditional salons were growing rapidly. International examples suggested that the model could work.
We conducted research, but it was not deep enough. We did not create the cheapest possible operational MVP before committing capital. We believed the product could take the market and started building.
Competition was not the main problem
Two competitors appeared with more than EUR 2 million in funding each. Our project was largely bootstrapped with approximately RUB 6 million. It would be easy to explain the outcome as a capital disadvantage, but that would be the wrong lesson.
The fundamental problem was that the economics did not work well enough. Customer acquisition, specialist travel time, schedule density, service consistency and repeat usage created a difficult equation. More investment could have delayed the conclusion without changing it.
We closed earlier than the better-funded competitors. That decision protected part of the capital and, more importantly, our reputation. A founder is not obliged to save every hypothesis. A founder is obliged to face evidence.
Twelve checks I now use before a serious launch
- Map the market: study direct, indirect and international competitors.
- Run customer discovery: verify the problem, frequency and willingness to pay.
- Model the full economics: include acquisition, operations, support, refunds, capital and founder time.
- Benchmark assumptions: compare your cost structure with public evidence from similar businesses.
- Build an operational MVP: use manual workflows and no-code before custom development.
- Sell personally: the founder should learn the objections before hiring a sales team.
- Start with constrained marketing: prove that demand exists before buying scale.
- Avoid one-channel dependence: acquisition that works only through one auction is fragile.
- Test seasonality and density: especially in local and on-demand services.
- Delay external capital when possible: evidence increases both valuation and negotiating power.
- Choose partners through demonstrated commitment: enthusiasm in a meeting is not ownership.
- Define stop conditions in advance: decide which evidence will make you close or pivot.
Closing can be a product decision
Founders often treat closure as a personal verdict. It is more useful to treat it as portfolio discipline. A failed model can release capital, attention and reputation for a better opportunity. The danger begins when sunk costs become an argument for additional spending.
A good stop decision needs three things: trusted data, a deadline and the emotional ability to separate the hypothesis from your identity.
This article adapts my earlier founder case published on vc.ru.