Keeping Track of a User
  • Suprabh Sanket

Keeping Track of a User

People from Ad-Tech Industry will agree that keeping track of a Customers/Audiences/Users is very vital, as it affects the revenue of an organization or ad-network.

This is because, if you’re displaying a relevant ad to a relevant audience, you’ll have better ROI as compared to broadcasting your Ad to the world.


Let us look into the various ways of tracking a user, across websites and applications. Also, how users’ info can be used effectively.

The whole intention of the system is to increase sales for an advertiser. To achieve good sales or conversion, the ads should be displayed to the most suitable users, associated with a product. They can only be identified if and only if we know enough information about them and their behavior, intention or interest about the product.


For example, if Mr. Harry, 45 years old, is working in a MNC then it is more logical to show him ads of formal shoes whereas his college going son might be interested in Running Shoes or a pair of Sneakers!


Then comes a question - how can we keep track of a user?

We can get user-related information from the user itself - i.e., through User-Generated content, where the user provides us information, like Name, Age, Gender, Qualification, Profession, Phone, Address, Interests and User IDs like G-Mail, Facebook, etc.


Secondly, tracking can be either or a combination of the followings -

  • Search Keywords – What are the Keywords a person uses in his/her online searches

  • Page Visit – What kind of Websites a person likes to visit or frequently visiting

  • Mouse Movement – Most clicked area on their device (by using Heatmap)

  • Ad Views – What type of Ad a person wishes to see

  • Clicks – After seeing an Ad, what kind of Ad a person is clicking

Or, thirdly, through the Cookies present in the user’s device. A cookie can be a First Party or Third Party.


> First-Party Cookie tracking refers to the dropping of a cookie in the user’s machine, against the same domain name that the user is visiting. This is useful in tracking the user’s mouse movement to find out the content or element, that attracted the user’s attention; and for how long did the user spent his/her time in the view.


For example, if Mr. Harry visited Nike.com. He spent some time to view running shoes, then headed to the buy button did not make any purchase. This shows that he was interested in the product, but for some reason, he did not buy it. Hence if we drop a Cookie in his machine and show him Nike’s ad on whatever site he visits, it’s more likely that he would buy in the coming days.


> Third-Party Cookie tracking, here all the user-related info is stored in the form of Cookies.


For instance, let ABC Inc. be an Ad-Network running campaign for Nike. And in their DSP, under Nike’s Campaign, they have whitelisted ten domains. If Mr. Harry visits any of these ten domains, a cookie will be generated and dropped (if not already present). The benefit of this type of cookie placement is that Mr. Harry can be tracked across multiple sites by the same Ad-Network.


After collecting info about the user, what can we do next?

Often websites do not have a “Register Page,” where users can go and enlist. Therefore, most of the users are not registered/authentic users. So, we should have a mechanism to track the user and their interests. For this, we can use -


> User segmentation - We can attribute segments to a particular user or group of users, based on the info we acquired by tracking cookies. For instance, a user can be cataloged into sections like age-range, profession, qualification, location, interest-based, device-type wise, etc.

To get more ideas about segmenting users and advantages, check out this blog.


> Recommender System - When all the data about users are acquired, using all the available resources and mechanisms. A hypothesis can be created, which can be used to make decisions required to make changes based on user performance. Furthermore, using the hypothesis and user’s data, the ad-network can create a Recommender System that can be used to recommend relevant ads to users.


Usability

The DSPs in the market nowadays generally build a user pool that contains all the necessary data about users, across its multiple clients and channels. This directory of User data happens to be organized in such a way that it can be utilized anywhere or by anyone using that DSP.


For example, in an Ad Network, a user data collected by Client-A can also be useful for Client-B (only after agreeing on some level of contracts between the parties, if they are interested).


So, there is no need to curate the same data every time a new Client or Campaign comes, let the Cookie do the magic!


Post Written By:

Samson Francis

Software Engineer

Kritter Software Technology Pvt. Ltd

35 views