Ask HN: How do you generally build a product?

Once I’ve proven, technically it works, its a slow iterative process of turning coal into a diamond.

Sometimes that process takes days, sometimes weeks and for me specifically, years. But it’s the drive to finish that actually gets you to a completed project. Drive and conviction that your thing is going to fulfill your goals (whether that’s money, scale, technical accomplishment, experimentation, etc).

2) Try to work on your idea at least 4 or 5 days of the week, no matter how small the time. You don't have to write code. Sometimes it could be something as small as choosing a color scheme, picking your twitter handle, registering a domain name. You just need to make progress.

3) Acknowledge upfront that this idea might not lead anywhere and might not be commercially viable but the knowledge gained is not lost and you can use that idea in something else. For example, I once worked on a pet project that involved turning SVGs into PNG or GIFs via Javascript. I also worked on another project where I tried to build a blogging platform on Google App Engine (I was trying to learn more about Flask and Google App Engine). Both did not go anywhere. However my latest project - https://nocommandline.com - uses the code/knowledge from those 2 projects. My blog is based on the blogging code and some of the images I use were generated from the other project. In fact I have used that in some other stuff I did for people

Gives you advantages: something you can share with others, a developed/implemented vision, something you can actually launch at any time without the underlying product for testing reasons, and something that will psychologically motivate you.

I used to build out backends first because it was easier for me, but many projects just didn't get launched. The landing page first is actually harder in my opinion, but psychologically having something presentable makes the project feel real.

https://medium.com/intrico-io/strategy-tool-amazons-pr-faq-7...

Not many ideas really. Usually its one or two. Maybe for a few months, years or several years.

And then I build them because I am a programmer.

The other option, for people who have money, is to create an inspiring vision, a mission, and a useful plan for accomplishing it. The use your large wallet to hire the best people you possibly can who really align with your mission. Make sure they stick to the vision and the plan. Re-iterate it if they deviate. Fire them if they really insist on going off in a different direction or if they don't perform. But at the same time, you need to be able to recognize when parts of the plan need to be updated. To avoid having to do that too often, its best to hire extremely talented people to maintain the lower-level plans.