Are startups 10x cheaper than ten years ago? | Hacker News Reader

Are startups 10x cheaper than ten years ago?

swombat.com

32 pointscfontes12 comments

I've been doing running internet/technology startups since '95. Here's my $0.02

* Labour costs roughy the same per hour. Although using the likes of elance or even mechanical turk, it is much easier to find cheaper labour for certain tasks.

* Hardware is a lot cheaper/faster. Strangely - due to the bloated & broken nature of things like Boost, compile times have stayed strangely constant! More importantly, the likes of VPS significantly improve the cash flow. A server to host your first website is now $10 a month with no contract. Ten years ago - you'd be locked into long contracts or buying a machine. You want twenty servers to stress test something, or test software on 8 different Linux versions, fire up some EC2 instances. There simple wasn't the equivalent back then.

* The development time to build most systems has got a lot less. The availability of libraries, cloud-services, APIs & virtual machines has drastically reduced the time taken to build certain types of systems. E.g. with something like Tropo you can quickly build fairly advanced telephony services. With AWS, you have access to huge computing resources should you need it. Web frameworks have significantly reduced the time to do CRUD-like tasks.

* The quality of FOSS is far higher now than it was then. Particularly the libraries and development tools. The quality of community sites (e.g. StackOverflow) also helps a lot in terms of access to development knowledge.

* Most markets are more mature now in terms of what is a minimum viable product. E.g. to build a web business now - it prob needs to be more polished then ten years ago.

* The Internet market place is a lot bigger. E.g. there are a lot more Internet users, app-buyers. Ten years ago - a business like AirBnB wold probably have not had enough users gain critical mass.

* The ease of reaching customers is easier. Things like Apple's App store (or even Facebook) makes it pretty easy to reach buyers, receive customer payments etc. Similarly the likes of twitter and viral/social marketing to all certain types of product to get a lot of cheap exposure.

* The ability to develop web based applications as opposed to desktop applications does significantly reduce development costs. Not that developing for IE6 isn't a pain, but it's a damn site easier than having to develop and support lots of different desktop operating systems and legacy versions.

Clearly the X variable is highly dependent on the space (as Daniel points out in the article). That said, the initial friction point for creating a startup is substantially lower than it was 10 years ago, especially for pure web based alternatives. Domain names, hosting, scalable cloud services, development/server licenses, support assistance, etc. have all dropped substantially over the last 10 years.

Today's startup might produce much more than yesterday's startup with the same funding, but today consumers also expect much more value for their buck than yesterday, too.

All startups are competing with all other startups. If it becomes easier for all startups to deliver more, of course, they must all deliver more to compete. But, as I pointed out, they also have many more opportunities for doing so (and many more competitors... it's a vicious cycle, isn't it? I think they call it "capitalism")

There is also a bigger market for the startups to work with, in the early internet days nobody gave a lot of importance to online services they were actually worst than local software, nowadays we can produce better quality software and better looking also so most of the users are willing to use (a few to pay) it in a desktop fashion. So now you can hit millions of users and earlier you couldn't, this helps online services to don't die faster then they can improve their service.

I was doing startups 10 years ago. You didn't need Oracle (first of all, MySQL was already available) and you didn't need tens of thousands of dollars of servers. Yes, you probably paid $400 a month for colocation but that's a relatively small amount compared to labor.

The market was smaller (fewer people online, less commerce) but it was also a lot less competitive. All things being equal I prefer the current situation.

I think even $400 a month was a very high hurdle.

Immediately it cuts out every individual who is unemployed or who is a resident of a poorer nation.

The fact that you can get started for $10 a month is an enormous improvement on how things used to be.

You didn't _have_ to colocate... you could get a shared server and pay a lot less, but give up root access. At the time though lots of sites were basically big CGI scripts and didn't really need their own server. Even these days you're still really talking $80-$100 a month for your own server, anything less and you're probably getting some kind of slice or VPS that you have to share with a bunch of other sites (as I was reminded this morning when Slicehost hard rebooted the machine my slice was on due to some other site misbehaving, taking down one of my sites temporarily.)

€49/m, 8GB, 2x750GB hdd, i7-920 quad-core: http://www.hetzner.de/hosting/produkte_rootserver/eq4

So, about 10x cheaper, in this case...

How is $80 10x cheaper than $400?

(Is anything X times cheaper than Y? Cheaper = delta. Expense / cost is the base. Seems like it's better to say 1/10 as expensive.)

Alright, 5x. I wasn't considering that the $ was so weak.

Same difference.

Startups in themselves have not become cheaper or more expensive. It just happens Moore's law has been almost equally nullified by the rate of growth of complexity of problems.

In a similar fashion just like computing was very costly back in the 80s but so were the problems that we could think of to solve using them.

Reply on news.ycombinator.com