Hitwise Intelligence - Heather Hopkins - UK
Analyst Weblog
« YouTube Takes 2 in 3 UK Visits to Video Sharing | Will Referrals from Search Engines to Retail Websites Reach 35% by Christmas? »
Internet Usage Data as a Measure of Offline Advertising Effectiveness - (Part 1 of 3)
September 27, 2006
"Half the money I spend on advertising is wasted; the trouble is I don't know which half" famously opined John Wanamaker. This quote is reportedly from the late 1800's - have things improved? Anyone who has spoken to me in the past six months, knows I have been focussed on the issue of using online usage data (search term volume and content and site visits) to improve the measurability of offline advertising and brand awareness. This week, we published a research report with Yahoo! Search Marketing and i-level (a digital agency in the UK) on this very topic. The research uses three case studies (Orange Shop, Sky and The AA to compare ad spend, creatives and online activity with online behaviour using Hitwise data on search terms, clickstream and visits.
Carlos Grande wrote a piece on the research in this morning's FT (registration required) talking about how internet advertising can boost traditional campaigns. You can request a complimentary of the research here and in a series of three blog posts, I will present some of the analysis from the research.
The key takeaways from the report are that:
- Online advertising is a highly effective medium for raising brand awareness and driving brand association;
- An integrated media approach can maximise the effectiveness of all media spend
- Online usage data, such as visits to websites, search term content and volume relating to a brand can be used for brand research and measurability of offline spend.
In this post, I will focus on the third. I will cover the other two in future posts in this series.
The first chart below is one of my favourite from the research. It shows spend by media for The AA Motor Insurance compared with searches for "aa car insurance". (The report has a similar chart for Sky, with similar results.)
The AA Motor Insurance spend is heavily weighted to awareness (television) advertising with 51% of the 2005/2006 budget going to awareness spend. Our analysis found that this spend has a very strong positive correlation with searches for "aa car insurance" and "aa insurance". Correlation does not equate to causation and car insurance is a highly seasonal product. How do we isolate the impact of spend on search behaviour?
By calculating the increase in searches for "aa insurance" and "aa car insurance" versus the increase in awareness spend advertisers can better understand the impact of ad spend on brand awareness. In this example, we found that from December 2005 (the month with the lowest total spend) to February 2006 (the month with the highest total spend), share of searches for "aa car insurance" increased three fold while total spend increased 11 fold. Television advertising made up the greatest increase from a base of zero in December to 57% of the total spend in February. Also, from January to February 2006, share of internet searches for "aa car insurance" increased 25% while TV spend increased 134% and total spend increased 86%.
A good way to isolate seasonal effects is to do year on year comparisons. The following chart compares share of UK internet searches for "car insurance" on the left axis with share of searches for "aa car insurance".
We found that searches for "aa car insurance" increased 58% year on year in February whilst searches for "car insurance" decreased 4%.
Using the Hitwise Search Term Suggestion tool, we found that there was a 70% increase in the share of UK searches that included The AA brand with the keywords "car insurance" comparing the four weeks to 25 February 2006 and 26 February 2005.
This is important as the generic term "car insurance" is highly competitive and relatively expensive in search engine sponsored listings. Last time I checked, the maximum bid for the term "car insurance" on the Yahoo! Search Marketing bid tool was £5.12. By contrast, the term "aa car insurance" fetched £2.09.
Hitwise search term data is updated weekly and so firms can get much more timely feedback on the effectiveness of campaigns. Marketers have traditionally relied on telephone surveys to measure brand awareness and brand association. Search term data is much more timely and because of this timeliness, allows marketers to better isolate effectiveness of particular campaigns and messaging.
We are open to working with marketers and academics who are interested in measuring the effectiveness of ad spend using online usage data. If you are interested, please post a comment. Also, as this is a new way to measure advertising effectiveness, we are open to comments and suggestions.
Part 2 in the series can be found here.
Posted by Heather Hopkins at 11:29 AM
| 
(3)
| 
(0)
TrackBack
TrackBack URL:
http://weblogs.hitwise.com/movabletype/mt-tb.cgi/330.
Comments
This is a very good articel, I would lov learn more about this topic.
Kenneth
Posted by Kenneth | March 22, 2007 11:50 PM
I find that internet advertising has the distinct advantage that you can directly monitor the return on your advertising dollars. By the traffic statistics, you can tell what's working and what isn't. This is directly opposed to a bilboard or tv ad where people could look at the ad and you'd never know whether or not it made a difference in their purchasing decisions.
Posted by Marketing Leads | May 15, 2007 10:59 AM
This research is highly interesting. First, it adds to the proof that online advertising serves a genuine purpose. It also adds to the idea that it can be used quite wisely and creatively in the pursue of a true business purpose: spend less money to generate more value. Thx for sharing the research conclusions with us.
Posted by François La Roche | October 13, 2006 03:46 AM