Monday, October 09, 2006

Using Google AdWords: a live experiment

Today we experiment on ourselves, like those doctors of old who swallowed harmful bacteria to test their own antibiotic creations. The topic is Google advertising. It can be a very cost-effective and productive part of a business' marketing mix. Or a waste of money. We'll see.

The first step was signing up for Google AdWords. Why Google and not Yahoo? We used both in my last company and consistently found that 65%+ of our hits were via Google. So if you want one online ad partner, Google is it.

The signup, including account creation, keyword selection, account verification and billing setup, took about twenty minutes. I chose about thirty keywords. Google has a nice tool that recommends keywords based on initial terms you enter. You want keywords that are as specific as you can create, and you can use multiple-word keywords (such as "MVNO startup").

My campaign focused around MVNOs. MVNO-related work is perfect for Google because the acronym is very specific and limited. Therefore, the competition for positioning isn't nearly as severe as it is for keywords like "personal injury," "insurance" or "mortgage."

You want keywords as specific as you can get them, so that only the people who might need your service will see your ad. Google ranks ads based on budget, so if you have limited-interest keywords, your ad will be positioned higher, cost less, and be more likely to result in a quality clickthrough (i.e., someone who's really interested in your service).

Then I picked a budget. I started with $30 per month. I can monitor the results in real-time, on Google's Ad Campaign web page.

To compare that expenditure with other marketing-related expenses, consider these statistics:

One year of Google AdWords: $360.
One trade show attendance: $1000-$3000.
Local chamber of commerce membership: $250.

Finally, I decided on some metrics. We'll look at them weekly. They are:

# of impressions (or times the ad is viewed)
# of clickthroughs (or times the ad is clicked on, bringing the user to my website)
# of customer contacts (times people fill in the "contact us" form on the website)
# of opportunities resulting from those contacts
# of sales related to the contacts--the bottom line result we want

The first two metrics come from Google. The third comes from my website. The last two I will track on my pipeline report.

So, the experiment starts now. We'll look at the first set of results next Monday. See you then if not before.

, ,