<< by on October 4th, 2010
The next session that Evan and I attended was the “Inner View: Google’s Keyword Research Tools” with Baris Gultekin from Google. After Baris’ brief presentation, there was an extensive Q&A session. The Q&A session was hot — lots of questions!
Baris is a Group Product Manager at AdWords. Baris gave an example of how difficult keyword research is. In August 2008, there were over 7M searches for “Britney Spears” globally, and over 1,000 various misspellings of “Britney Spears”, and the top ten of those misspellings received an additional 915,000 searches!!
Some facts about keywords:
- 55% of quereis use more than three words
- 70% have no exact match results
- 20% of queries on Google have not been seen in the last 90 days
Google has many tools you can use for keyword research, so which will work best for your needs?
- Search-based keyword tool
- Traffic estimator
- Insights for search
- Keyword Tool
- Wonder Wheel
The Keyword Tool has over 2M users that get 300M+ keyword ideas per month! The newly released tool is designed to give you both more power and more control via filters. Some new usability features are coming in the next few days based on user feedback.
Why did you get rid of the old tool?
The old infrastructure was going away, and they felt the new version would be more powerful for the end user and more keywords as well. The audience’s general sentiment was that they did not like the new keyword tool as much as the old one. Basically, everyone wants the old tool back!
Baris said that much of what seems to be missing from the new tool is going to be added back in the system in the next few days. So the new tool should be feature-similar to the old tool very soon.
How can I get more exact match search volume?
Best option is to use the Traffic Estimator tool.
When will Geotargeting be a part of the keyword tool? (Evan’s question!)
It’s in the works to get the more granular data down to city and state level.
There are two global monthly search volume is number of global searches, independent of your language or country selections, and local search volume is specific to your language and country selection.
What is the minimum volume a keyword needs to show up in the keyword tool? What percentage of keywords searched in Google never make it into the tool?
There is no volume threshold per se, but there is a privacy threshold. So any query can get into the tool’s results, so long as it doesn’t jeopardize privacy. The keyword focus is on commercial words, versus those that might be more private searches.
They try to balance efficiency and information. So they target the words that will cover most of the topic.
Should you be logged in to get the best results?
If you are not logged in to Google when using the tool, you are limited to 100 results. You’ll get even less if you’re not logged in and using filters.
Are you channeling us towards the Google Suggest with this tool?
The backends used for the Google Suggest and Google keywords tool are the same. Google Suggest uses searches from Google.com, and Google keyword tool uses those queries and related ones.
What’s the best way to find common misspellings, since this tool doesn’t show them?
The new tool will allow you to have more control, likely also allowing you to choose to see misspellings. They haven’t done this earlier because broad match captures this.
What’s the timeframe for the new tool?
Over the next year.
Do you pull in the realtime data that could spike numbers?
The numbers are averaged out for each month, and the download has an average for each month for the last twelve months.
Many of the numbers are exactly the same. Why?
The tool rounds the numbers.
Why does Facebook get filtered out? Doesn’t Google like Facebook terms? For instance, a keyword tool search on “Facebook” yields “Facebook login” first.
This may be a bug. It is possible that there are no ads on the term “facebook”, and that would mean there were no clicks on an ad on that search, which is what drives the numbers in the Keyword Tool. Or, if the click was obtained, but the quality score was low, it also may not show in the keyword tool.
Marty Weintraub pointed out that this is limits users to focus on commercial searches.
Is the tool using impression data where there was an AD impression against that word?
Baris said they track every search. However, one audience member said that this ignores the long tail. For instance, if you sell “easter baskets”, the data is heavily skewed towards March/April. So is Google showing the volume for the keyword because it had an ad display sometime within the last twelve months? Baris said yes.
Do you consider ATT and AT&T the same word?
You will currently see two separate stats in the Keyword Tool.
Is what you get via the API the same as from the Tool?
Not really, but there’s a bug fix that is currently fixing it.
What is Broad Match like in the tool? Just like AdWords?
For instance, if you search “flowers”, you would not get “roses” with the tool, like you would with expanded broad match in Google AdWords.