Mahalo, arguably the most sexy Mediawiki design out there, is a new vertical search engine. A humanly edited search engine.
When online mediahype boy Calacanis set out to start a new search engine, this was the result of an old personal feud with SEO.
Months later Mahalo has seen the daylight and after some weeks, I must admit that Mahalo could impress. But does the service really stand a chance or is this another A-list hyped platform?
As Google fan, I must admit that last years I have always found what I was looking for and who will tell me what the best result is. Other than that best result a purely subjective factor is.
Who delivers me the best search result? Social platforms such as del.icio.us where I can see how many times geeks have bookmarked certain sites and not other ones. Or is the best result that blog entry written by a guy who has only 300 page views every month?
So far I see two major problems with Mahalo, and in analyzing I obviously kept in mind that Mahalo still in alpha stage is. If an operational, public site really alpha can be, unless alpha the label is to apply when your wiki hasn’t enough of entries yet.
The top argument to use a humanly edited search engine is at the same time its weakest factor. Human editors make the results slow, slow in update rate. Who guarantees me that Mahalo will be able to keep up with the ever changing internet content and deliver up to date results.
Together with human editors comes bias. And bias can influence search results, especially when I look for commercial results. Just have a look at the results for wedding rings. Who guarantees me that the top 2 results, shops really are the best results? And how could I make it in to the top 7 if I had a wedding ring shop? Offer Calacanis some dollars or a ring?
The excessive Google implementation can be a burden and degrade the quality of results, sometimes even be the prove that Mahalo’s human factor is contra effective. Best example of this is the news page. Although Mahalo offers many outlets for actual news, if every time I am confronted with Google News at #1, I might as well just use Google News.
Other times I am happy that at least there are some Google results implemented on the site. Even though I know that Mahalo’s aim is to manually write the results for the Top 10000 search terms, when I need a search engine, it also happens to me to search further than the Top 10k search terms. And when I google, I will most of time dig deeper and deeper, using more and more narrowed down search queries.
Mahalo does the opposite for non existing search queries and offers related entries, broadening the results again. What is a better example than a search for Calacanis SEO? Although the first Google result didn’t exactly lead me to Jason’s rant, it had a link to the page I was looking for. Mahalo results are non existing, and related results are everything except related in this example.
The sad thing is that both problems send me back to Google. And that can’t be the purpose of a new search engine, unless I miss the broader picture.
Originally posted on July 8, 2007 @ 12:45 pm