Targeted Online Advertising – An Analysis

Over the past few centuries, there have been many efforts to come up with methods to help predict human behavior — what Leon Wieseltier of The New Republic calls mathematizing the subjective.

In the digital era today, with rapid and disruptive technology at the forefront of innovation, the internet has become an important tool for almost every individual and every business. In this digital age, with connectivity of multiple devices, we have taken on multiple sets of identities,one being the physical self and the second being the digital or cyber self. This has given immense scope to marketers to extend their direct marketing campaigns and reach out to customers more likely to purchase their products. This has thus spawned a number of content hosting companies and third party content providers to fight for this highly lucrative share of the $23.4 billion online advertising industry (as of 2008), in a study conducted by the Network Advertising Initiative (NAI).

Targeted online advertisements work on the mechanics where cookies of each website track the movements of the user across cyberspace, storing data on parameters such as clicks, time on webpage, and path of movement across cyberspace etc. This data is captured and stored by many data providers who build data bases on user’s identities and then try and make an estimate of what the person browsing is like in real life by comparing the data with the existing set of data. This data is then given to advertisers who bid to be able to showcase their ads to the viewers. This, in a nutshell is the business model behind targeted online advertisements.

These advertisements fall into four broad categories, those based on Demographics, Psychographics, Behavioural Variables and 2nd order activities, such as online activity, website history or contextual advertising. The avenue of Behavioural variable based advertising was observed to be the most common method to advertise online.

As far as the efficacy of these advertisements go, there is no question that they are successful. The sheer ubiquity of these advertisements is testament to the lucrative business model that it is. Further analysis of the NAI study shows that advertisers pay up to three times more for cookie based ads and that these ads secured 2.7 times the revenue as compared to non-targeted ads. Futher, the study also highlights that these ads were twice as efficient in securing converts of clicks to buys. This, therefore is conclusive evidence that targeted advertisement is an effective way to reach out to customers. Most users though, still consider these ads as an intrusion of their privacy and are increasingly worried.

Add 1

The challenge however, for marketers is not from a financial sense. The economics of the targeted advertising are quite sound. The marginal costs of targeted campaigns are far lower than brand related searches ($1.69 per search versus $15.65 per search). The marginal cost of a click is only 13 cents versus that of 72 cents in a non-targeted campaign. Third party data providers, through their sophisticated algorithms, have the ability to gauge the user’s age, marital status, whether the person has children, income level, education level and a motley of other data. The scary prospect of this is that the user’s anonymity is increasingly in name only and can have many negative connotations. As an illustrative example, banks can use this type of data to get apriori knowledge about a person’s income, habits, gender, and race and ascertain credit worthiness in unsavory manners and not issue loans/credit cards solely on the basis of credit worthiness in the traditional sense. The prospect of artificial intelligence getting more and more ‘intelligent’ has the plot of a fabulous science fiction movie written all over it!

This ineluctable flood of targeted advertising requires stringent monitoring from the federal agencies if there is to be a semblance of order and transparency to prevail. The increasing data available online of many people who unwittingly give data away has the potential for large scale exploitation and misuse of data. It thus falls to both Federal agencies and companies to play their parts right. If only more companies like Google practiced the motto, ‘Don’t be Evil’, cyberspace would be a much better place. That said, targeted advertising is a great way for marketers to send specific relevant ads and also reduce their advertising budgets. If regulated properly, it can be a win-win situation for both consumers and companies. However, as with any other tool or technology, data analytics has its own set of limitations and thus should not be overused or depended upon, illustrated by the image below.


Add 2

Adarsh Pande

The above article is a winning article for Sanrachna (the article writing competition organized by Marketing Club, IIM Bangalore) by Adarsh Pande, 1st year PGDM student at IIM Ranchi.

IIM Ranchi

This handle is managed by the Media and PR cell of IIM Ranchi. You can connect to them via mail at


One comment

IIM Ranchi

The article is pretty lucid and comprehensive ! No wonder why it’s a winning article 🙂