These aggregated measurements only represent part of the campaign result.
More insightful data such as user's feedback, preference, amount of time spent, demographics, online behaviour, and purchase intent can all be derived from various server logs. Unlike traditional media, these parameters can only be extracted and processed in an online environment, hence online campaigns should utilise this information to the maximum extent. The web server logs detail data collected for each request of a web page. Examining these logs often reveals valuable information, such as the time-of-day when the majority of requests are being made. In the scenario of a fastfood campaign, the click through rate might increase progressively as the time-of-day approaches lunch, and hit an apex around X hour before lunch. Utilising this knowledge, the campaign can be optimised to concentrate more advertisement inventory in the hour to X hour before lunch.
With click-stream processing, the information in the web server logs can be rendered into users' sessions. Vital information such as behavioural interests and intent can be data mined from these sessions. For instance, a user session may include a sequence of activities: review PGA news, search for 'Tiger Woods', check the weekend weather at Pebble Beach, and compare the price of the latest fairway wood. From the above sequence, the user indicates behavioural interest in golf, plus intent to purchase golf equipment. With this data, a sports equipment company can schedule golf-equipment banners to precisely target users interested in golf.
Since online behaviour is derived from specific user activities, they can be made as niche as required. For example, users with travel interests can be further segmented into those intending to purchase a safari trip to Africa and those interested in Egyptian pyramids. In a travel agency scenario, the first group is ready to spend, so they are targeted with special safari offers via direct marketing. Meanwhile, the second group might be directed to the company resource centre for information on Egypt.
Instead of promoting mere awareness, the firm is now capable of catering specific products and services to a corresponding segment of online users.
After all, the web encompasses more niche market segments than any other traditional medium. And behavioral targeting opens up vast opportunities to reach these user segments. In conclusion, a tremendous amount of useful data is buried in some form of logs on web servers.
With the process of data mining and a bit of creativity, the data can become vital information for an online campaign. And most online players have already packaged data-related processes and analysis into their online marketing arsenal.