MyLiveSearch Not As Completely Useless As I Expected

I’ve been beating up on MyLiveSearch for months now. The startup promises to create the first real time search engine, where results are indexed literally as the search is conducted.

It all started back in May when the unknown startup got big press in Australia as a search startup that “Google is keeping a close eye on” (with no supporting evidence whatsoever). WebProNews called the service “revolutionary” based on a phone conversation with one of the founders and without actually testing it. The company, which sports a 1995ish design, also missed its much touted June launch date. All of this hype provided lots of ammunition for some easy trashing.

But now the startup is set to launch this Wednesday, and we’ve had a chance to test it over the weekend. While I don’t think the company has much of a chance of displacing Google as the king of search, they are innovating in the area of real time searches. And their approach to search is unique in that all the work is placed on the user’s computer.

First, MyLiveSearch is not actually indexing the web. Users must install a browser plugin (Windows/IE only) and then select a base search engine to start every search (Google, Yahoo, or Microsoft). Users can also select a different starting point, such as Wikipedia.

The first ten results for a query are pulled from the selected search engine or other starting point and displayed. Then MyLiveSearch begins to spider out from those sites and index other results, which are displayed as they are generated. All of the spidering is done from the user’s computer, though, not MyLiveSearch’s servers. This is the reason for the plugin requirement, of course. And it has the side benefit of putting basically zero resource strain on MyLiveSearch itself.

Overall the results are not super impressive. Google is indexing many websites nearly constantly now, so real time results are less important than they were a couple of years ago. As an aside, Google’s faster indexing time has also affected the blog search engines, which are really only useful today to show chronological or reverse chronological results – Google is indexing many blogs as fast or faster than they are.

But MyLiveSearch is attacking the search problem in a new way by pushing all of the work to the user’s computer. Users still get Google or Yahoo (or whatever they like) data for the top ten results, and then see real time stuff spidered from there.

There are still a number of bugs and annoyances with the service that need to be ironed out. Only one search can be conducted at a time and you can’t have multiple windows open using the service. You also can’t refresh a results page, forcing you to re-enter the query.

MyLiveSearch won’t be the most likely to succeed search engine to launch in 2007, but it probably won’t be the worst, either.