Saturday, September 27, 2008

Google Contextual Advertising - an example

Let's now have a look at what you can do by expanding your Google Adwords advertising a little further afield - through advertising on Google's content network.

Here's my results for www.ogcg.com.au.


Search Advertising

For a start there is an even distribution of searching across the OB and GYN groups with a slight increase in the PR group. The Cost/Conversion rate is also better in the GYN group compared to the OB group - perhaps less competition so cheaper advertising. Also note the incredibly small click thru rate for the PR group. There may be several reasons for this: the text to the Adwords ad is very specific about seeing a medical professional, the ad might not be well targetted or people may be after something specific that they don't believe the website would provide. The cost per conversion is also terrible: 7-14 times worse than the GYN and OB groups.

Perhaps it is not worth advertising using the PR group terms (or not all the terms in the group).

Let's compare the Ads delivered via Google Search to those delivered through Google's content network - ie across websites that have AdSense advertising.

Content Advertising

Contextual Advertising (placement) enables you to have your ad displayed on other people's websites who use Google's Adsense system. Google claims your ad may appear on millions of pages across the web.

Visitor Analysis

Looking at our data you can see a very different distribution of terms with contextual advertising - in the case of the GYN and PR groups the ad generates between 2 and 6 times the impressions that Search advertising does. This shows that not everyone uses Search all the time.

Quite often people are visiting a website eg perhaps if they are pregnant they might already know which site they want to go to and spend a lot of time there. I'm sure the same is true of most areas of medicine and the web in general - people seek information first but also a sense of community and a website is a collection of information and people around the theme of that website. This however doesn't necessarily mean they are more likely to purchase something.

Click Thru Rate (CTR)

Notice next how the CTR is much reduced. This is probably because people were doing something else at the time. Who looks at Ads when they are in the right place? Also perhaps for a content rich site the ad may be displayed on several pages a visitor clicks through while on the site. After first noticing the ad it is probably likely to be ignored on subsequently page visits.

Cost Per Conversion

Here is a strange one though: the cost per conversion is much lower. Why would this be? I think this is primarily influenced by the price I put on the keywords in the first place. I pay a flat rate for contextual ads but often have to pay more for search advertising. Perhaps there is also less competition meaning you are charged less.

Placement on specific sites vs general contextual advertising

Finally, you can choose to either place your ads on particular sites or let Google place it across all pages where your terms are relevant. Looking at our results you will see that never have the sites chosen delivered better results than the untargetted contextual advertising even when we are getting many more visits.

Why is it so? I suspect the placement algorithm used by Google puts your ads across a much more relevant group of pages than just picking certain websites.

Is Contextual Advertising worth it?

Overall I'd have to say we are pleased with the results from contextual advertising - it has enabled us to reach potential customers we could not reach before at a cheaper cost than search advertising.

Remember that periodically you need to perform your own analysis of your data, make hypotheses and test and change as a result of what you learn. Think of it like investing - just splashing money around is unlikely to achieve optimum results.

There are several limitations of this analysis. Feel free to point them out.

People who read this post also read :



No comments:

Post a Comment