The business of pastry sales saturates pretty quickly relative to other products. So "better pastries" can keep on scaling but you can't keep charging more for them. The regional food/dining marketplace can only bear so much, so "better pastries" can't draw from a much larger group of customers.
There's nothing inherent in software that makes it so that can't be the case for software products either. I anticipate the response here is that some products are more complex than others -- true enough, but how many people buy a wedding cake at the grocery store?
My wife and I did in fact buy our wedding cake from the Publix Bakery near the church. We were extremely satisfied with the quality and the much more affordable price. Enough cake for a couple hundred guests for about $400.
To expand my point, we had a really tight wedding budget but still wanted to have friends and family. We had our reception in the Church's Fellowship Hall, not some fancy venue that cost 10x more. My wife and I had to make some hard decisions (harder for her than me, I could have gone much cheaper), but in the end we stuck with the plan. Are more expensive designer cakes worth the money, absolutely! And the pastry chefs who make them are worth their wages. We met with one of the local bakeries that won on Buddy the Cake Boss. But! We didn't have the money. No amount of wishing could make us have the money at the time. We loved our wedding cake from grocery store bakery. It was yummy with good icing and our guests said nice things. Maybe its just because you always say nice things at weddings.
In software, startups can be the same way. Do they need the super scalable, secure, best in breed software? Sure! That'd be great. But sometimes you just have to get out the door now and do improvements later.
My wife and I are just as married as we would be if we had an expensive cake. There are startups in business because they started with what they could afford and then got traction.
