Google Trends Flaw

So, there’s been a lot of chatter going on about Google Press Day and the new product announcements. I especially like this entry by That Girl From Marketing

Google releases a bunch of beta products… hours later “Me Firsters” rush to test and write about said products. One week later Search Engine Optimization tool makers incorporate aspect of these products into SEO tools. Two weeks later spam will show up in one or more products…

Three months later non-SEO Bloggers write about said products as if they discovered them. Four to six months later Google releases APIs of said tools. Six months later start-up companies incorporate said tools/APIs into product offering, call themselves Web 2.0 and then are covered by the earlier “Me First” new sources…

Nine months later the same “Me Firsters” will write about: a) how some aspect of the tools suck b) how the tools are collecting info for some larger conspiracy / silly named update c) how they’ve seen aspects of said tools show up in the regular Search Engine Result Pages or d) a retrospective article on said tools rehashing their earlier posts with links to other posts they’ve written on the subject. A year later Google releases more beta products and the cycle repeats itself….

Funny stuff… because it’s so true! Except, I am going to skip ahead to the nine month scenario…

I was at Google’s office in Newport Beach yesterday to eat lunch with our rep and check out their new digs. They only had one section completed, but it was very cool. It was burger day… good stuff. We rapped about a bunch of random things… from upcoming movie releases to the speculation going on about Yahoo and MSN. Anyway, I mentioned a flaw I noticed in the Google Trends tool and our rep has since confirmed it. Hopefully, I can explain this correctly… it’s a bit confusing…

While I was playing around with various keywords, I found that Google Trends is not delivering queries as exact matches. For example, it appears that a search for “Tiger Woods” also counts as a search for “tiger.” So, the trend for “tiger” will always be higher than the trend for “Tiger Woods.” Try the query yourself…

Now, it might make sense that more people would search for “tiger” than “Tiger Woods.” However, If we check Overture’s Keyword Selector Tool, we see conflicting information:

tiger vs Tiger Woods

(By the way… not knowing what the results would be, I used Tiger Woods as an example because he was working out at the gym below Google’s office while I was there.)

So, let’s try a couple more queries to prove my hypothesis…

“spears” vs “britney spears”

“angeles” vs “los angeles”

“eyed” vs “black eyed peas”

See? Who searches for “spears” or “angeles” or “eyed”…??? The multi-word phrases should have higher search trends than the single words, but since “britney spears” counts as a search for “spears,” the single word will always be on top.

Of course, Google Trends is in Google Labs… so it’s basically in beta. I admit it’s an entertaining tool to use and the integration of news and Cities/Regions/Languages breakdowns are interesting, but I wouldn’t count on it just yet. Just like a lot of the other tools out there, it should be used as a relative measurement.

Update: Nice find by Phreaki… By adding the minus sign (“-“) and the word you want to exclude from broad matching, the trends appear to be more accurate. See comments…

Post to Twitter

Subscribe to SEOdisco

Subscribe to SEOdisco by RSS feed RSS    Subscribe to SEOdisco by email email

Digg Delicious
Reddit Technorati
Stumble Socializer

Related Posts

    Fatal error: Call to undefined function related_posts() in /home/seodisco/public_html/wp-content/themes/notso_freshe/single.php on line 54