Behaviour trumps Demographics

One of the earliest things I learned in my career was that behaviour online was not a good indicator of who the customer was. In 2000, the company’s marketing was focused on reaching customers through targeting communities and queries, paid and organic, that either directly or indirectly related to the product. As a pure web company, without any traditional marketing expertise, demographics were never considered.

Segmentation by demographics can be important when media reach and consumption is limited by time and geography. Traditionally in broadcast media intent is hard to quantify, leaving demographic information as the best basis for good decision making. Online, there are more tools to use in identifying the audience in a more meaningful way.

There are a large number of variables which can indicate intent and interest online, and a number of ways these can be used to reach and communicate with potential customers:

• Behaviour on site
• Online community participation
• Search query selection
• Inbound traffic source
• Upstream & downstream activity

Behaviour data can also help to find the value in traditional demographics as well.

The broad generalisations implicit in demographic classification of traffic are problematic for a number of reasons. Online media fragmentation and differing skill levels mean diversity of experience within previously homogeneous demographic groupings can be very high. The efficiencies that demographics created in identifying and communicating with targeted markets within mass media are no longer a concern on the internet. With programmatic solutions that allow for finer targeting of small groups within a site, or through a network, breaking the audience up into large, general groups is not much easier than catering to smaller, more focused audiences.