Web Databases & Applications

During the dot-com boom, I used an online database creation and publishing service called Bitlocker to create a bug tracker for my start-up. At the time, we had 20 people using our Bitlocker based bug tracker spread over the US and India.

Bitlocker was an unique product at the time. It allowed you to create a collaborative web-based application without a lot of programming. They provided templates for applications like CRM, bug tracking, recipe management, employee review tracking, etc. It was sort of like FileMaker Pro for the web for free!

While Bitlocker was a WYSIWYG application, it exposed the underlying schema to the end-user and so you did need to know a little bit about databases to program with it. Also, modifying the structure or the UI was not simple and led to all sorts of data inconsistency issues.

The service eventually failed for the same reason many other dot-com companies died. It wasn't profitable.

Today, I think there are two interesting alternatives to Bitlocker.

First, there is Oracle's HTML DB. HTML DB was originally Oracle's answer to Microsoft Access. But in reality, it is so much more. You can use it to build complex and scalable applications very quickly. It does require a fair amount programming ability (SQL & PL/SQL) to do anything complex. BTW., it is free if you own an Oracle database.

Next there is JotSpot. JotSpot is very cool and simple. It is a web-based custom Wiki builder. You don't have to really think about data when you start building a JotSpot based application. It is ideal for collaborative tracking applications that involve unstructured data. BTW, it requires no programming skills.

The Long Tail - Overcoming the 80/20 Rule

Venture Blog has a very interesting note on the 80/20 rule based on an article from Wired.

The basic premise is that the 80/20 rule is dictated by physical constraints. Those constraints are removed when we move to alternate models enabled by the web. For example, Barnes & Nobles carries around 130,000 books based on what they believe will be the top selling books. On the other hand, 50% of Amazon's revenue comes books ranked lower than the top 130,000 books. Amazon does not have the same inventory constraints that Barnes & Nobles has and thus can carry many more books. Most of these don't sell many copies but each small sale adds up. This is the long tail.

The same can apply to business intelligence. The trend today is to provide users with pre-built reports. There are only so many pre-built reports that can be provided economically by an IT organization. On the other hand, BI tools such as Oracle Discoverer have evolved so that they are simple to learn and use while at the same time can be delivered over the web and managed centrally. I believe that we are at a point, where IT organizations should really consider empowering their end-users with more capabilities and thus serving the long tail.