Sunday, February 21, 2010

28th Amendment. Anyone on board?

Congress shall make no law that applies to the citizens of the United States that does not apply equally to the Senators or Representatives, and Congress shall make no law that applies to the Senators or Representatives that does not apply equally to the citizens of the United States .

I got this from my CPA Don Brooks (who is a pretty cool guy.)

Tuesday, February 16, 2010

Perils of 65% market share

We tend to see market leaders in two dimensions. We see how much ahead they are, and we see how high they have climbed. Its hard to see the third dimension behind this facade. What is that little object in their mirror?

Today I am trying to find that little object. Right now its in the rearview mirror of my favorite search engine.

According to BBC news, the market-leader-du-jour: Google has consolidated its position and now enjoys a safe 65% market share.

The number 65 has magical powers. It repeats.

Back in the early nineties when Novell was king, they enjoyed 65% market share. Back then, their hold on it seemed infinite. I remember driving to Provo Utah to meet some engineers at Novell. It was like a pilgrimage of the NLM developers.

And while they were at the top of their game, Microsoft --at the time a puny DOS company-- launched a thumb pin into the heart of Novell's business. And kept pressing. The rest is history.

Anyway, now that google has hit the magic number, I wonder whose next. It can be Microsoft. Cannot be Yahoo. Cannot be Facebook. Cannot be Ask. What if its a non search company?

More on it later...may be not.



Thursday, February 4, 2010

Commodity Services: Does size matter?

The mundane matter of selling commodity services is usually on my mind these days. It has almost become a favorite pastime. Partly because of the challenge and partly because I myself own a few companies who sell commodity services every day.

Today's question: Does size make a difference in the way you sell commodity services? I do realize that ADP uses different tactics to sell vs. my local accountant/payroll service. But is size itself a sell-able feature? Does it work both ways? --Is being small also an advantage as being big?

Here is a list of personal and professional services I personally use. And my preference for size of business:

Haircut - Small, local, somewhere i have had a good experience.
Personal Taxes - Small, local, somewhere i have had a good experience.
Coffee - If its as good as Starbucks, I prefer a small non franchise store. Otherwise Starbucks.
Food - Small, local, non-franchise.
Legal - I wouldn't buy commodity legal services if possible.

Here are some business services:

Payroll - I am with a big name payroll company but I shouldn't be. Competence issues.
Accounting - I am with a local CPA.
Temp Agency: I am with Robert Half / Accountemps. But only because I saw their billboard. And because I believe their pricing is standard and constant regardless of the customer.

Anyway, back to the size issue. Here is a trial recipe:

Local Presence in a non franchise way. (IBM of New Haven vs. IBM Global)
Known names and faces (locally)
The image and personality of a small boutique shop.
The stability and performance of a big enterprise.

Conclusion: If size of a service company is to be used as a feature, smaller is better.

I need to experiment with this.


Monday, February 1, 2010

Is there a cloud for my software?

Our new SAAS product Hyperconversion is going into beta soon. It helps people to Create, Manage and Host Landing pages. We've been at it for a while now and its a beautiful implementation of the point-and-click school of thought.

I have an interesting dilemma. Was wondering if someone reading this can give me an idea or two.

The system is developed in ROR and designed with a share-nothing architecture. Each client gets his own snapshot of the full application. We designed it this way so that we could run thousands of applications on Amazon AWS or Rackspace cloud.

Here is the problem: Amazon does not update the image of a running virtual machine. So if the image crashes, we lose everything. When we figured this out, we found RackspaceCloud.com which is a lot cheaper and solves the running image problem. But the problem with rackspace is that they don't have elastic IPs. If a server image goes down, and you end up creating a new one, you get a randomly assigned IP address. There is no way to get a block of IPs.

We need persistent IPs because our system depends on DNS to client's domain names to the correct landing pages.

Does anyone out there know of a cloud solution which leases IP blocks and also saves the running images?

Will appreciate your thoughts. We cannot run dedicated servers. It has to be a cloud based solution.