What is the source for these statistics. I do not see a supporting link?

Todd, great point.

The research is mine - but I can certainly give you the raw data. A little setup first.

http://www.WordTracker.com is the default in the Internet marketing community for determining search strategies, trends, etc. They (pay for) and grab enough sample searches to be viable accordingly. They ain’t cheap - but they’re worth every dime.

Hacks in this space will tell you to go to Overture’s search (now Yahoo) tools to figure out the same because it’s free. Problem is it’s only a sampling from Overture - which isn’t viable.

But the new changes a Yahoo killed their free tool. The lads from Wordtracker just launched http://freekeywords.wordtracker.com/ - as a sampling for their larger database (brilliant move by the way). I of course had to play with it and got my number from there.

Take a look - but warning - lots of fascinating info - even in their free tools - you could get lost for hours.

Paul

And then I turn to KeywordDiscovery.com (another popular tool used by professionals in the field) and get a 597 count for “podcast how to”, 300 count for “how to podcast” and no results at all for “parrot podcast”.

Further, “Battlestar Galactica Podcast” barely registers and “video podcast” is in the hundreds.

Not sure what that means, but there you go.

Fascinating.

I’ve always preferred Wordtracker in my research. It’s seems like the majority of users do to (from what I’ve heard/read). Of course, popularity has never been a requirement for excellence.

One thing I do know is that Wordtracker only keeps for a 3 month timeframe where KD holds onto stuff for a year. That might explain the differences too.

Fascinating. Fun with numbers.

Paul

Under the category of FWIW…

Researching keyword phrases is always tricky and understanding the bias of the system you’re using is critical.

I don’t want to get into too much detail, but here’s some factors to ponder that affect results.

- wordtracker is the last 90-days. BSG runs mini seasons so when a season is “in play” it’s reasonable to assume a spike in searches.
- keyworddiscovery is the last 12 months so can provide some insight into longer term usage.

- wordtracker is mostly based on results from actual searches on dogpile.com and metacrawler.com. Users of meta search engines tend to be more internet saavy (aka. geeky) than say an average yahoo searcher. BSG being scifi with a big geek fan base… well, you fill in the rest.
- keyworddiscovery says they include results from dogpile.com and metacrawler.com yet show almost none of the same results, it is unclear how long it takes for data to roll into their database. I’ve seen many quirky results from them in my years using the service.
- keyworddiscovery does has a huge skew in their global database including results from deleted domains (the kind with the ads and searches on them), lot’s of pay-per-click ad engines and the results can be strange.

If I had to choose one research tool, wordtracker generally has the cleanest results that, in my experience, translate into real world searches the best.

I use keyworddiscovery to help find anomalies and see if anything strange shows up over a 12 month range.