Commenting the ordinary

Today Hank Williams wrote a post on his blog that I can relate to. Starting Internet companies has become more expensive over the last years according to Hank. He argues that even with open source frameworks and on demand cloud computing becoming cheaper and more linear in their cost impact, the main cost driver for building and operating still is human resources.

I agree: having a team of bright developers is still the prerequisite to building a competitive web application/service. This has not changed over the last years. Even offshoring in my experience does not reduce the cost significantly. For me offshoring is a necessity to get access to required technical resources attached with a lot of risk and overhead to identify and manage the right people you want to work with.

One development in the Internet economy is responsible for making the launch of a new Internet service more expensive: There is a lot of competition for the users attention. Countless companies coming up with services that are thrown at users who have to try to comprehend what each service is about and whether it generates value or not. To be on top of this Startup Noise is a challenge. You either have to buy yourself in front of your audience – too expensive for startups and arguably not an efficient use of capital – or you have to find other ways to virally get your product spread.

But why do we see much more Internet startups these days than just a couple of years ago? In my opinion to get a working product out of the door is much easier and cheaper than it used to be 8 years ago. And there are a lot of people that have the skills and have been part of an Internet culture for some years where it seems to be natural to at some point in your life try to launch your own product/application.

So while I do not completely agree with Hank that starting a new service is as expensive as it used to be (otherwise we would not see so many new startups) I do agree that getting your startup in front of your audience is more challenging which makes it ultimately more challenging to start a successful Internet company.

Google launched their Google Health service yesterday, that allows me to build my medical profile online in one central location. I guess I would be able to connect to my doctor and have a more efficient exchange of information. This might even allow doctors to ‘connect the dots’ across different diagnoses from different doctors, creating in the end a more complete picture and hence improved health care for me as the patient.

I am by nature wary about providing to much information about myself in one central place (probably related to my German/French background with its inherit distrust to governments and other centralized institutions). Therefore it struck me when I read on Fred Wilson’s blog a post where he wrote about his motivation to make is medical profile public (though Google does not allow for this).

He received critical comments back outlining the problem of insurances rejecting certain people based on their existing medical conditions or medical profile which might place applicants in a high-risk group. Now insurance companies try to make their decision on accepting to new policy holders based on information disclosed by the patient. The patient might hold back some information with the risk that this might endanger his coverage in the future (if the insurance company finds out). This whole game of information disclosure upon application is an interesting game under asymmetric information. The question is, would it be more efficient for the patient and/or insurance companies if all parties had perfect information?

As long as health care is organized around private profit maximizing insurance companies, perfect information would allow health insurance companies to optimize their risk portfolio leaving high-risk patients without insurance. Therefore providing perfect information in conjunction with private insurance companies is not Pareto-Efficient and does not increase aggregate welfare. However perfect information would lead to a better, Pareto-Efficient market in case of a universal insurance scheme, which would be either public or public regulated (There are however inefficiencies ingrained with public service schemes – the question is if these are offset by the overall gain in public welfare).

The launch of Google Health and the discussion around centralized and even public health record are interesting in the light of the presidential election and the plans of the two Democratic candidates. Depending on who wins the elections in the fall Google’s service could take off.