Keyword search numbers you might find interesting

[thecableguy]

Here's an example of how Google's keyword tool can skew the search numbers, Wordtracker/Seobook does the same thing as well. The following is some of the search numbers that common sense will tell you that it can't be right (yes it's exact match searches). These are just some REALLY obvious examples.


mothers day gifts 27,000
mothers day cards 14,800
mothers day flowers 9,900
valentines day gifts 18,000
easter baskets 12,100
electric heaters 8,100
skis 18,100
snow shoes 6,600

The reason is that Google/Wordtracker/Seobook goes back 12 months to get the monthly average of the search volume (12 months total divided by 12 = search volume). A look at the "Local Trends Column" (click on "Columns" and check mark it if it's not displayed, should be by default) to get a better picture.

Some months have a huge spike in searches, while other months drop off to virtually nothing. But the search numbers for the previous 12 months average out to the numbers that you're seeing. Click on "About this data ?", then "Local Monthly Searches".

For evergreen seasonal niches in the long run probably won't make a difference, but for items that have a shorter shelf life it can, like electronics or HDTVs. Depending on where in the "Trend" you are the search numbers may be skewed.

[John Williamson]

Precisely. Great examples.

[dburk]

Hi thecableguy,

While you offer some good advice about the need to check for trends caused by seasonality and other factors, I disagree with your characterization of these results as "skewed". The tool seems to report exactly what they claim to report without skewing.

Perhaps we should say the data "accurately" reports the average for the past 12 months and isn't skewed for trends, so you need to look for and consider how a varying trend might affect your campaign.

[thecableguy]

Originally Posted by dburk

Hi thecableguy,

While you offer some good advice about the need to check for trends caused by seasonality and other factors, I disagree with your characterization of these results as "skewed". The tool seems to report exactly what they claim to report without skewing.

Perhaps we should say the data "accurately" reports the average for the past 12 months and isn't skewed for trends, so you need to look for and consider how a varying trend might affect your campaign.

Yeah I guess I kinda "skewed" that one

What I meant was most people assume that it's like the old keyword tool that gave the search numbers for the previous month. And don't take into account the local trends and the 12 month average when calculating the search count. My bad. A lot of people skew the data and cut the search numbers to close.