Categories

Think about it

Build it

Run it

Market it

Measure it

Uncategorised

The attribution error: Issues in aligning the results of online and traditional marketing Wednesday , November 18 , 2009 by Hasnain Zaheer

Save to My Web Business

The attribution error: Issues in aligning the results of online and traditional marketing

  Hasnain Zaheer explains the problem of attribution in advertising and presents a brief review of initiatives and solutions with a view to more accurately measure and track online advertising and line up the different channels to their conversion.     In traditional advertising, there is a separation between brand advertising and direct response advertising.

Brand advertising seek to change consumers’ attitudes towards the brand and orient them to favourably consider the brand. Stop. It does not measure any conversion.

Direct response advertising is the method that elicits and measures response. It typically relies upon coded coupons and tactics such as inviting consumer responses and orders in to special telephone numbers mapped to campaigns.

Traditional advertising relied upon this separation and had no problem attributing sales. It simply attributed all sales to the direct channels.

Online is a different beast. For one, it is possible to measure any online advertising served through any online channel, in any format - direct, brand, widget, video, text, banner, pay per click, pay per view.

Cookie is used to measure conversions. The problem is the new cookies overwrite the previous cookies. As a result, the last cookie wins and all the previous online advertising that the customer may have been exposed to, are ignored.

An environment of ‘Last Cookie Wins’ skews the results in favour of direct response oriented ads. Even within a single channel, paid search, it skews the result as it gives credit to the last search query typed (perhaps a navigational keyword to reach the company’s Web site) not the one which attracted them to the Web site when they were comparing products.

As an example, let’s assume we are running:  
  1. Banners on MediaSmart – Sensis’ media network 
  2. Video advertising to promote the brand on Fairfax Digital Media 
  3. Paid search campaign on Google AdWords and Yahoo! Search Marketing for keywords that are organised into groups to target different stages of buying cycle
  4. Listings on shopping comparison engines e.g., Shopping.com 
  5. Content network campaign on Yahoo Search Marketing Paid 
  6. placement ads on Google AdWords 
  7. Affiliate campaign with several dealers running as affiliates of our products on Commission Junction.
  The above represent a fraction of advertising channels available to an online marketer today.   A consumer may enter our advertising environment or touch our brand from one, a few, many or all of these channels but will use only one to purchase at the end. Add traditional media and advertising to the above concoction and the picture gets even more murkier. A purchaser may have taken many offline and online actions before the final purchase.   A few possible scenarios of consumer interaction are:  
  1. Saw a billboard and formed a positive image of the brand
  2. Saw a TV ad and reinforced the image
  3. Typed a search query and clicked a paid listing to reach your Web site to check out the specifications and download a whitepaper
  4. Visited a showroom to check how it looks like
  5. Then compared prices online and clicked on a link to buy it from a dealer offering the lowest price.
  Most products cannot be sold at the first advertising exposure. Consumers must be made ready for the purchase and sometimes it may take many (five as in the above example) exposures before a consumer buys.   
So how does the marketer know what proportion of the final credit should be attributed to each advertising touch-point. Which creative is working and pulling more than its weight? Which creative is slowing down the process? Which ad is accelerating it?  

Example: A TV ad made lots of people to search for us but our Web site was poorly search optimised and the paid search campaign didn’t target the right keywords. So how do we attribute credit until we can measure at each step and attribute credit (or take it away) at each of those stages.

Now, that we know what is attribution and how it is ‘mostly in error’ we can have a look at some current initiatives:    
  1. Ad servers are designed so that they serve creative to a user in a certain sequence. The first timer is shown an introductory banner, a specified widget when he is exposed second time and so on. There is no attribution, only an attempt to move the user to a goal. The downside is, how is the advertiser to know if the chosen path will indeed lead to success.
  2. Behavioural advertising firms can put your ad in front of the people who are most likely to respond. Others can test your ads and put the best performing ads in front of people.
  3. Surveys are presented to users who were exposed to online ads at different Web sites and compared to a control group of users who were not exposed. The questions in the survey attempt to measure brand awareness, brand favourability, purchase intent etc.
  4. Customer satisfaction research and customer’s or respondent’s propensity to recommend the product also provide us measures that can be mapped to online and offline advertising programmes.
  5. Clearsaleing technology integrates leads generated from online marketing initiatives to a CRM.
  6. Pay-per call advertising in which the final action takes place with a call can be traced back to the keyword in a search campaign.
  7. Service providers such as Australian Avanser that allows a large number of phone numbers to be allotted to discrete pieces of online or offline advertising and can calculate conversion arising from those conversions.
 
The above initiatives cannot help us attribute each unit sold to the exact set of specific pieces of advertising – creatives, keywords, channels – that the customer met on his way to the sale. However, the above initiatives and solutions help travel a part of the way.  

We are still probably in stone-age as far as attribution solutions are concerned. In any age, we would never enter the mind of the customer to determine what offline factors affected the customer most.

But, initiatives to more accurately measure and track online advertising and line them all up to the conversion are underway and it would be one of the most exciting fields to watch in online advertising in the next few years.  

Have I missed anything in this discussion on attribution in advertising? Let me and others know in your comments.


Rate this article


Leave a comment











  Make me a member in WebBusinessAge


Anonymous Feb 15,2010

<>.. ] is another must read source on this subject<>..]

Anonymous Jan 26,2010

The information here is great. I will invite my friends here. Thanks

Anonymous Dec 21,2009

I somehow dont agree with a few things, but its great anyways.

Anonymous Dec 13,2009

Very great website. The information here is very important. I will give it to my friends. Cheers

Anonymous Dec 04,2009

I am definitely bookmarking this page and sharing it with my friends. :)


Login to Web Business Age

Web Business Age Community

News Letter


Tags

advertising marketing engagement conversation paradigm pull b2b marketing b2b magazine b2b email retail e tail blue ocean strategy cheenu srinivasan IBM IIBV media businessweek marketingprofs mckinsey quarterly long tail chris anderson seanx sean cummings google small business startup online business strategy entrepreneur digital entrepreneur enquiro global agency digitas copyblogger copywriting web business eclipse australia digital ministry 10e20 design dropshipping drop ship copy social media twitter facebook meetingmix newscred attention economy apple web writing maria veloso web copy blog themes corporate blogging blogging business blogging blog design blog usability blog marketing blog business corporate corporate blog blog platform avanser brand advertising attribution attribution management clearsaleing direct response branding attribution error behavioural advertising pay per call direct marketing digital strategy web analytics online marketing internet marketing web metrics finding fertile ground scott shane technology based venture technology enterprise tech venture analytics performance kpi measurement website metrics pageviews unique visitors visits tracking display advertising ppc seo optimisation link building paid search search engine on page off page website marketing basics marketing strategy web strategy online business positioning market segmentation jack trout al ries Philip kotler marketing management strategy web design usability customer web design user interface accessibility web interface interface design ui web developer full service digital agency choosing implementing website domain name Australian domain name indian domain name in au domain premium domains choose domain name

Advertisements