Hacker News Reader

Comment by phkahler | Hacker News Reader

I think this is going to be important now and in the future. Systems are getting so complex nobody can build something complete from scratch. I see people making high quality components and then merging or being acquired to create a whole end user product. Prior to that point, you'd have something that's not very profitable unless lots of companies license your component and do integration themselves.

View on HN

So, an interesting thing I observed while building a SaaS company was that you can actually get off the ground with a very simple solution, you'll just end up serving smaller customers with relatively simple needs (say $50-100/mo.) These initial folks will pepper you with feature requests while still paying you money for what you've got, which can give you feedback about which complexities you're missing that you really need to implement. If you choose which things to implement carefully, you end up making your product more useful to larger customers and you'll start closing some larger deals. (E.g. $500/mo.) Rinse and repeat and before too long you've got a product with all the bells and whistles and you're closing enterprise deals. (E.g. $2,500+/mo.) The numbers here are all relative, but the point is that the smaller customers end up subsidizing the development of what becomes an enterprise product.

(I think the advantage this approach has over the one you've suggested is that the product is always capable of making money even from the very beginning, just on a smaller scale at first. I think this is a critical feedback mechanism so that folks know whether what they're developing is actually useful to folks.)

Great comment. In a way, it's like what Geoffrey Moore talks about in his book "Crossing the Chasm", about the difference between early adopters and mainstream users.

Systems are getting so complex nobody can build something complete from scratch.

Come out to Microconf sometime. You can meet ~300 nobodies all at once!

"Systems are getting so complex nobody can build something complete from scratch."

I would claim there is still a large chunk of areas that have very poor service from the software field and would find tremendous value of a trivial software product that is just applied correctly. So it's more about service design rather than hightech, if one is just looking to make a profit.

Please forgive me if this presumissious of me, but you wouldn't happen to know of some of these areas that need more quality software? The only one that I'm aware of its medical, which usually has regulations and long leads before making money.

I'm mostly clueless. But I see the pattern all around me still that in established business processes there are so many places where lots of things are done as in pre digital ages. B

As an example...

Building and construction is still in infancy as what it comes to utilizing digital technology. The field is fragmented with long subcontractor chains. The digitalization that happened in e.g. car manufacturing is not yet complete. The field is dominated with clunky, expensive cad packages which, while some may be eatabslished standards, are never the less decades old piles of legacy software. Most of the data is still transmitted between parties in paper drawings.

Reply on news.ycombinator.com