HN user
Can't say generally as I don't have data, but there would be too many live threads to follow from people that the viewer may not care about. Unless there is some level of automated moderation of live threads and possibly some linking to "friends/people I care" it could quickly become noise on the page. Just my2C. But I still think its a nice idea.
Truthfully said. Thanks for trying to get the expectations real. Your post makes me think further that we are not even considering the emerging market opportunities wrt e-commerce. Most people in developing nations don't have credit cards or easy means of doing e-commerce, but its definitely getting there. Lots of scope, and entrepreneurs in the valley should definitely also focus on these markets and get in early with market specific products.
Nice post. I agree to many of the upsides of using AWS or for that matter any other webservice, but down side is that, cutting over to your own scaled system later would be difficult. Also, when a system scales there is a lot of money saved by efficient engineering (initially costly though). Paying per-use fee is not economically scalable as it adds significantly to the Opex. Keep in mind that the best monetization vehicle for most Web2.0 technologies is online advertising; which pays per access and is little, so such a per-use Opex will never let the profits scale.
For early-startup and prototyping AWS is definitely a good option.
Not seen such a service yet, seems like a nice thought. Shouldn't be too difficult to implement though (Streaming data options abound), not sure how much market will be there.
Yes ofcourse! The struggle is ongoing; especially when you have a team of software developers. We do continuous market research by regular interviews/polls, keeping up with news in the interest area, talking to people (or forums) dedicated to your target area and finally looking at current companies in the focus area. We have one engineer give 30-40% time on market research and other's keep tab on specific focus areas.
Keep coding on the side, work on the re-usable components (in case you change the focus area or the main value add). Example, infrastructure (databases, user management etc), maybe getting your garage ready for primetime. Even if you discover the idea you have would not work in the market next idea can have a quick ride on this infrastructure.
And finally, experimenting is the best way to find out. If you are thinking revolutionary then any type of market research will not be conclusive. If the idea is evolutionary then you have lots to gain from existing market research (get membership of some marketing companies) and you own!
Best of Luck ~M