Over the last 15 years, I built a unicorn and sold it for more than a billion and half. We wrote unit tests extensively.
I'd be curious to learn more about your experience, would you be open to a chat on sqwok.im? I'm somewhere in the middle with moderate unit tests and e2e for critical paths.
The idea that you need to throw unit testing out to be successful is kind of silly.
When you’re growing a very complex system (huge micro services pattern and dealing with sensitive data) you’ll be happy you have tests to save your ass.
…speaking from experience.
I mean, who would even think it is?
There are also alternative approaches to achieving quality, and there was that report/"study" circulating around that claimed Team Software Process actually beats most agile/extreme programming methodologies in at least one metric (or two: defect rate and time to completion, IIRC). Of course, none of the actual "agile" development teams do clear-cut TDD or Extreme Programming either, so you are never comparing apples-to-apples.
As a corollary, I'd also say that neither does the UI or UX matter that much to success, and there are simply a bunch of intangibles that can make or break a product.
But one "tangible" thing is that you've got to be serving some customers' needs, and need to continue to do that. And rare are businesses which have a need of "my button needs to be rounded and 10px away from the next button" or "my code needs to be unit tested": those are actually "my customer can get their work done through my product" and "it should rarely break".
None of this means that you should not invest in pretty/functional UI/UX or well-tested code. Just don't expect that to mean much in the grand scheme of things. They will affect your ability to hire top-level people once you are an established business, though.
(I personally am too attached to "high-quality" [as in well tested] code and would struggle to kick off a start-up for that reason only, even though ideas keep popping up :)
Product/market fit has more to do with functionality and the minimum quality design aesthetic required by the customer.
In our case, we disrupted a rather boring, traditional field. Our UI/UX was never great neither did we hire “frontend” people until post-acquisition but performance and data security were huge to our clients thus testing.