[zuppy]

it's more like short term gain vs long term gain. experienced engineers can design an architecture that will allow you to scale cheaper and faster in the future, at the high initial cost. it will be cheaper to maintain, better for security.

depends at what point your business is at the moment of hiring and what you plan to do with the product. do you need volume or quality (both variants are right)?

[throwway120385]

If your business is going to cease to exist in 4 months, who cares about scalability? Pay the interest when it comes due and when you can afford it. If someone is serious about building a company they will be okay with that.

[swiftcoder]

Yes, this is a very important aspect. An early stage startup needs zero-to-one engineers. People who build fast, aren't afraid to break things, and don't mind YOLO'ing a year of their career on a gamble.

If you find product/market fit before you run out of money... that's when you need to hire engineers who are in it for the long hall. People who focus on reliability and scaling. People who might stick around for 5 years to see if your startup becomes a unicorn.

[gedy]

Sure but then incentivize engineers to hack it out knowing they'll have to deal with the shit show if you become successful. Sorry but most "startup engineers" aren't , and it's basically bad for their careers to implement "the vision" in a throw-away manner.

[binary132]

unpopular opinion with engineers but unfortunately true

startups are generally moreso a business endeavor than an engineering one, although the engineering must correctly support the business

the engineering begins to take the driver’s seat as the tech debt and cost of scaling catch up to successful companies and begin to create excess drag

but for many years, such companies can typically still afford to throw away money to solve business problems, including these problems of scale

[rrr_oh_man]

Counterpoint: Experienced engineers will design the architecture that is appropriate for the current state of the business.

[jerf]

You know, three years ago I would have said that I can give you a pretty good architecture fairly quickly but if you just want banged-out code I'll be beaten by someone who just plows forward for at least a couple of months... but after some vibe coding I've done I think I could do both at the same time now fairly well. Vibe code very quickly that I also know I can make scale fairly well with not much more effort.