Data Targeting in Markets Without Third-Party Data

We live in an age where data is being used to influence business decisions. Conversations around Big Data are becoming more frequent while the volume of data being collected and analyzed by companies is increasing at an overwhelming rate.

To make most of the data available, in markets like the U.S., there are organized data segments that can help drive ads straight to very specific audiences. However, in Asia the data market is still quite nascent, but we are on our way and unlikely to ever turn back.

When we talk about data influencing online marketing, one of the more obvious manifestations is the use of the cookie (cookies are tiny files that sit on a user’s browser allowing the owner of the cookie to recall contact with that user’s browser). Cookies have their pros and cons and we have no intention to predict the future of the cookie here in this article. The point to note is that today, cookies are the main mechanism for employing data for targeting online display ads. By using cookies, a marketer can see what contact the user’s browser has had with their website or other online assets. This can give an idea about where in the consideration funnel the said user is, and what next steps could be taken to further improve the engagement.

The available data can be segmented into two clear buckets. Information that is specific to a particular website (that is issuing the cookie) and its users’ cookies are referred to as first-party data. Data that belongs to another person that could be used use to increase the efficiency and accuracy of ad buys is referred to as third-party data. As mentioned earlier, we are in very early stages in the Southeast Asian market in terms of access to third-party data units.

Although third-party data sets are not yet widely traded in Southeast Asia, data exchanges are emerging to fill these gaps, and the purpose of these exchanges is to support the open trading of third-party data “segments” that act as critical ingredients in a successful trading recipe. Some typical basic data segments that help enhance success at the impression level include:

  • Cookie data arranged by demography (age, gender, etc).
  • Cookie data arranged by interest in content (business, soccer, parenting).
  • Cookie data arranged by intention (product research behavior).

The first two are pretty straightforward examples. If you sell cosmetics, your efficiency would be significantly raised by excluding men (using demography data) to enhance targeting. This is not the same as targeting female-oriented content (that may still have significant male audiences) this is about targeting individual users known to be female. But are they really female? There will always be doubt and methodologies will range in accuracy and transparency but the best test for using elementary data sets is utility; does using it create performance uplift? In the absence of other details, performance stands as the best guide.

The ways in which these segments are created vary, but the data will either be declared or inferred. Declared data is information that the user has given to the company collecting that data. People share their details with retailers, social media sites, service providers like ISPs and banks, and a host of other companies. This data is sometimes sold to third parties, hence the rapid emergence of data exchanges over the years.

Inferred data, on the other hand, is created by observation of behavior. This can be a simple analysis of media consumption or exhibited retail research behaviors (the latter is referred to as Intender data.) User behaviors signal their intention to act/engage/buy and those identified as positive intenders become immediately more valuable to marketers. It is then up to the marketer to determine how best to engage the intenders in order to move them down the path to conversion. This is where customization becomes critical and data decisions far outweigh buying clout.

Given the limited availability of third-party data in Southeast Asia, how can marketers make the most of it? I have three suggestions.

1. Think small: Traded media is about blending high-yield/low-reach strategies with high-reach/lower-yield strategies. Remarketing is an easy way to start. Try not to outsource; keep your data.

2. Don’t throw anything away: Tag everything. Get a DMP to store the data. Create your own segments; test their utility. Compare your data with the emerging broad data sets from companies like Eyeota, Experian, and MediaQuark, etc.

3. Be patient: The industry is working on it. More data will come. Cut your teeth on remarketing and other low hanging fruit first. Your new skills from being creative with the little that is there, will come in handy, when you have a wider menu of options you could be buying.

The rationale for the above suggestions:

This area is moving fast. The number of companies in the space is increasing and more companies are focusing on the problems associated with the data market. As the RTB and programmatic spends increase (we see video GRP buying to be a significant factor in this increase) the market for data will grow, too. Publishers will work out that there is a significant benefit to selling data and earning revenue on ads that are not served on their pages but on their audiences.

We definitely see a very bright future for data marketing (particularly inferred data). However, we need to take these baby steps and importantly, keep a keen eye on the best interests of the public. Respect and privacy is a key ingredient in the success of data marketing.

At the current rate of progress, we think that within the next two years, Asia will be a leading market for data marketing and might possibly be the world’s leading player.

Related reading

An image of a pie chart styled as a medal, with stats about video content underneath.
A series of green arrows pointing into the distance, one reading NO ADS in white block capitals.