With the recent launch of Jonga, and Funnel having been around a few months now, it’s been an interesting time for local search engines. I thought I’d do a quick comparison. My highly unscientific methodology involved choosing a site I know, entering a search term I think should find it, and seeing how closely the results matched. I awarded 3 points for finding the correct page as the top result, 2 points for returning it elsewhere on the first page, 1 point for an indirect link, and 0 for anything else.
First off, looking for the Noordhoek Environment Action Group, I entered Noordhoek Environment as search terms. Jonga was off to a disappointing start, returning mostly guest lodges and property, and an outdated article about the Kaolin mine. No sign of the NEAG website. Funnel was much better, returning it as the top link. Google did the same, in fact doing even better for a more relevant link following that. Ananzi had NEAG 5th on the page, with the usual range of holiday accomodation up front.
Jonga: 0, Funnel: 3, Google: 3, Ananzi: 2
Next, I chose the Ethical Co-op, and entered the terms Ethical Co-op. Not as easy as it looks, as the site is new, certainly hasn’t been optimised for the search engines, and has hardly any links leading to it. This time Jonga came up trumps, returning it as the top result, while Funnel couldn’t find it, but did find an indirect link to it from my blog. Google also returned it as then top result, while Ananzi returned nothing of relevance. Jonga: 3, Funnel: 1, Google: 3, Ananzi: 0
Next, I decided to get sneaky, and choose a blog that’s South African in the sense that the author lives here, but which does not have the .za giveaway – Cosmic Seriosity Balance. Entering Cosmic Seriosity, unsurprisingly, none of the sites returned the desired result. Somewhat surprisingly, Jonga and Funnel had my blog as the top result, as it contains a links to the correct page. Google returned other blogs with the correct link, while Ananzi provided an interesting diversion with articles about cosmic vacuum cleaners, but nothing of relevance.
Jonga: 1, Funnel: 1, Google: 1, Ananzi: 0
With Jonga and Funnel neck and neck, I changed tack for the fourth test. Looking for an IOL article I remember as one of the most popular of all time, entitled KZN cannibal caught in the act, I used IOL cannibal as the search terms. Jonga first looked like it had got it completely wrong, returning a top link mentioning the Athens Olympic Games. But somehow its actually a story about a cannible, probably wrongly tagged. Close, but not what I’m looking for. It only provided two IOL stories before diverging (my blog again :*( ). The top article did provide the correct link as a related article though, so Jonga scores a 1. Funnel disappoints, with an article about endangered species as top result, and not even a secondary link to the correct article. Google again reads my mind, returning the desired article as the top result. Ananzi tries hard, with some good guesses, but not even a secondary link to the right article.
Jonga: 1, Funnel: 0, Google: 3, Ananzi: 0
Finally, I chose a short-lived link. Bidorbuy was featuring an auction on a Rosenthal Glass Vase. I tried Rosenthal Glass Vase auction. Jonga bombed out entirely, with no results. Funnel tried hard with a link to another auction site, and a vase (not Rosenthal glass, whatever that is). The phenomonal Google got it spot on, returning it as the top result. I’m starting to suspect Google is reading my browser cache 🙂 And Ananzi didn’t come close.
Jonga: 0, Funnel: 0, Google: 3, Ananzi: 0
With Google so far ahead, I decided to stop here. The scores:
Jonga: 5, Funnel: 5, Google: 13, Ananzi: 2
The results are clear enough for me. Jonga and Funnel are both vast improvements on the dated Ananzi, but fall short against the imposing standards of Google.
I don’t know how you think this is scientific. I can only hope that you are joking. It’s not really a true or fair reflection of the effectiveness of any of the “compared” search engines.
Comparing search engines is not something that is easily done, and certainly not something you can do on a rainy afternoon.
Perhaps spend a bit of time brushing up on your google fu?
What? A sample of 5 sucked out of my thumb not scientific? Say it aint so. 🙂
Nice little thought provoker.
These quick ‘studies’ are often quite useful as this is probably how the majority of search engine users think. They are not aware of the scientific accuracy, or algorithmic perfection of the search engines. The vast majority of internet users work subjectively, just like this test.
Thanks for taking the time and contributing to the discussion.
Note to previous posters: read the introduction for the “highly unscientific methodology”.
As far as the IOL Cannibal story goes I think explanation is quite simple why neither Jonga nor Funnel picks it up. Google would’ve spidered the story when it was first published on IOL’s Front Page, but because there is no other local reference/link to it (except yours in this blog now), neither of the tow new 2 search engines would’ve picked it up. They should now though. Ananzi also probably missed it because it doesn’t crawl enough.
This is why linking and sitemaps are important aspects of maintaining a healthy site.
As far as Funnel’s and Jonga’s algo’s go, time will test whether they deliver relevant results, gain usership and make them commercially viable. Google.co.za certainly does not deliver very relevant result a lot of the time. Another factor will also how the search engines classify search and blog spam and weight that in their SERPs. Google should theorectically have the advantage there, having been around for so long and refining their algo’s all the time.
You can’t compete with Google and win
It comes down to computing power that you have available, you can’t fit 150,000+ computers into your garage. Another issue is storage and bandwidth.
I found that Funnel has my pages 5 months old in their cache
Jonga has 2 months old pages
Your relevant search is based on outdated data stored in each search engine