Pronet Advertising

Google's Dice is Useless, What I Learned in 20 Clicks

Not long after the news spread that StumbleUpon was to be acquired by eBay, Google announced what many considered to be a direct competitor to StumbleUpon. Today I learned once again that this is far from the truth.

Before we were going to report on the new Google offering, Matt and myself were debating the merits of Google's implementation of their web-discovery tool was and really doubted the usefulness of basing the results on web history.

The problem is that many people don't use Google to find stuff that is 'interesting' to them, but they conduct work-related/school-related searches.

I decided to further test this out. I just enabled my Google Web History, installed the toolbar and added the dice button and decided to see what Google could offer me. Here are the first 20 results I got, why I think I got them, and why they are completely irrelevant.

Result 1: Chicago Transit Authority - Because I used their website to check what times the buses run.
Result 2: Text Link Ads - Because I was researching their site for an article I was writing.
Result 3: Geo TV - Because I searched for an interview of Musharraf
Result 4: Metra Rail - Because after I found out the bus schedule I also needed to find the train schedule after that.
Result 5: Google PageRank Prediction - Because I write for Pronet?
Result 6: SEO Tools - Future PageRank - Because I write for Pronet?
Result 7: Chicago Graduate School of Business - Because I go to The University of Chicago
Result 8: Add Page Rank Checker to your web site! -Because I write for Pronet?
Result 9: Chicago Craiglist - Because I was looking at what countries and cities Craiglist is available for.
Result 10: Search Engine Strategies - Because I write for Pronet?
Result 11: Ping-O-Matic - Because I use the site to ping services.
Result 12: Chicago MSI - Because I wanted to check what day the museum didn't charge an entry fee.
Result 13: Keyword Selector Tool - Because I write for Pronet?
Result 14: Northwestern University - Because I go to The University of Chicago.
Result 15: Build your Site Map online - Because I reported on automated Sitemap discovery.
Result 16: Chicago Weather Underground - Because in Chicago you never know what the weather is going to be like in another 15 minutes.
Result 17: Pakistan International Airlines - Because I'm going to be flying home for the summer.
Result 18: Google Rankings - Because I write for Pronet.
Result 19: The Seminary Co-Op Bookstore - Because I am a part of a Co-Op and buy books for college from there.
Result 20: Check PageRank - Because I write for Pronet?

Based on the above-mentioned 20 results it would almost seem as if Google is oblivious to the taxonomy of web search! According the the paper, there are three main reasons why people search:

1. Navigational - The immediate intent is to reach a particular site.

2. Informational - The intent is to acquire some information assumed to be present on one or more web pages.

3. Transactional - The intent is to perform some web-mediated activity.

With StumbleUpon, Me.dium, Streaker, LeapTag, and Google Dice, the space is getting crowded very fast. And given the current state of each of these tools, it's not hard to see why StumbleUpon is number one.

Think Search Engines //, April 30, 2007

“Analysis of Google’s Dice - Useless” — Analysis of Google’s Dice - Useless by Pronet Advertising Muhammad Saleem. I decided to further test this out. I just enabled my Google Web History, installed the toolbar and added the dice button and decided to see what Google could offer me....

julien, April 28, 2007

not only is the party getting crowded, it's getting crowded with people who misdirect you when you ask where the bathroom is.

Mike Bogo, April 28, 2007

Google's tool is just a cheap shot at StumbleUpon - I feel like since they lost out on the bidding war, some manager argues: "If we can't own it, let's beat it!"

The only problem is that their dataset and implementation aren't right for this kind of thing. And that's also why StumbleUpon's advertising campaigns have such low ROI.

Craig, April 28, 2007

I found the same thing. Although I like the Web History function, I am yet to find anything interesting in Interesting Items.

I removed the Dice from my toolbar after about the same time as you. Mike's right. It was retaliation almost. However, I think Google could head the pack at some stage.

As for Queries Related to your Search. That's completely useless.

eBusiness, April 29, 2007

I concur I removed it as fast as I installed it. The results were 'targeted,' but not high-quality new discoveries like StumbleUpon provides. Funny, the poor results of their toolbar were just as disappointing as recent results in their search results.

http://www.ebizmba.com

Asia, April 29, 2007

I wasn't impressed with the new feature enough to install it, mostly out of fear that Google will hijack my settings and throw my SERPS into chaos again. However, I installed the Google Personalized Page Recommendations tab, which is quite disappointing. Many of my searches are immediate needs and not something I wish to rehash the next day. I search for code constantly, but do not want to look at websites about code (boring!) - send me an obscure website on art because I love it and I will love this feature as well.

Apoorv, April 30, 2007

I installed the Personalized recommendations tab a few days back and my initial experience has been very bad. The recommendations are irrelevant and useless. Removed the tab as soon as i blogged about it.

Matt Keegan, May 3, 2007

I am beginning to see a pattern here...

Thanks for the feedback, although I knew that I wasn't going to bother with the Stumble Upon competition. I have found (and read) this site several times over the past few weeks simply by stumbling. I am amazed at how well targeted S.U. is, so there is no need for me to use some other tool.

Knowing Google, they'll work hard to overhaul and perfect the tool. In the meantime S.U. will probably make their usual improvements and stay one step ahead of Google.

Let's hope so!

adam, May 4, 2007

It's neat, no it's not as great as StumbleUpon...but the thing that Im wondering...is yes you use the dice to discover new pages.

When the dice takes you to a page, this page whether you deem it a worthwhile recommendation or not; gets recorded to your "Web History". Now, from here, that recommended site that the dice showed you, will be further affecting your recommendations "down the road" since you visited it in the first place.

Google needs to do something i think, base the recommendations on how long the user actually stayed on the site (cause they obviously wouldn't wanna copy SU with a "I dont like this" button lol), so say the user was only on the site 5 seconds (cause they've already seen the site, know what it offers, or maybe they just dont care for that result) that particular "dice roll" shouldn't affect further recommendations.

Katinka Hesselink, May 4, 2007

I do think this service could be improved by some sort of tag 'I like this'. I mean, not every site I search and look at, is worth saving.

Anything merely based on my search history is going to be limited at best. The genious of stumbleupon is precisely that it collects stuff people think is worth sharing. Only about 10% of the sites I stumbleupon are given a thumb up by me - which means only they get recommended to people like me, and influence my future stumbles. Same with thumbs down. And why would google not let us say: I looked at this site, and did not like it? I mean, part of my webhistory alright, but it was totally useless (or something).

BONTB, May 9, 2007

Not bad Muhamed, i think Google Dice is not worth either.

micky, May 11, 2007

yay - i came here googling for the Google dice URL to get the link to write the same thing in my blog. i shouldn't have bothered.

TimToTango, June 22, 2007

Hi Muhammad!

Nice story, I had not compared this google dice thing yet with the more 'social' networks like Digg and Del.Ici.Ous.

I think in the future I will do some experiments as well on such things, sounds nice. It will fit right in with my ther internet experiments, like the one I currently run on The Incredible Badger about Google Hot Trends (and which doesn't really seam to work :-)

Good luck,

TimToTango