In Part 1 I touched on the issues Internet publishers and big data industries face with competition and targeting the right content to the right people at the right time, in this Internet age. Before I talk about Youneeq and its feature set as the Predictive Analytics solution (will be detailed in Part 3), I’d like to go over some more detail about the pros and cons of rules based behavioural content targeting vs. predictive analytics algorithms.
Under rules based behavioural content targeting, content is delivered according to predefined rules. Visitors are assigned a unique cookie ID that tracks them throughout their web journey; the platform then makes a rules-based decision about what content to serve to the visitor. The rules tend to be founded on generalities such as demographics, location and the past activity of visitors with similar traits.
This system works well with smaller niche sites that implement specific content, and it has its place working alongside more robust systems incorporating complicated behavioural analysis algorithms. But rules based behavioural targeting systems are limited in function. These limitations are as follows:
1. Manual Maintenance – These are manual based systems with the inability to learn. Web administration and marketing need to work to greater lengths monitoring, defining and implementing rules, changing them on a regular basis based on assumptions.
2. Not Enough Power – Since rules based systems are based on assumptions, providing visitors with content relating to demographics, age groups and locations is too general and not powerful enough to deliver more relevant experiences to each unique visitor. Because of these limitations, return on investment potential is not being maximized and ad campaigns fall short. There are countless individuals with different interests and backgrounds that need to be taken into consideration.
With predictive analytics algorithms, visitors are targeted with very specific content. Unique profiles based on individual data are created by complex mathematical algorithms that monitor a visitor’s behaviour, and then deliver relevant content and ads to that visitor. These systems have the ability to mine data across sites and in some cases across networks, predicting what the individual visitor is interested in with near pin point precision. For content rich sites, a well built predictive analytics solution is detrimental. Here are the advantages of predictive analytics algorithms over rules based systems:
1. Fully Automated – Administrators and marketers do not need to manually build predefined rules models based on general behaviours. These systems incorporate advanced mathematical algorithms that monitor and learn about visitors.
2. Increased Accuracy – Since predictive algorithms learn about individual visitors, they can accurately predict and pinpoint content to serve individual visitors based on their actual behaviour. Over time, this becomes greater as the system works with complex evolving behaviours, building and adapting unique profiles about individuals.
3. Real Time Delivery – Large amounts of data can be analyzed, parsed and delivered to visitors in real time, delivering relevant content immediately across networks.
Part 3 will be posted later this week.
