Website personalization (aka web personalization) is an increasingly common online marketing technique for increasing the relevancy of the content by adapting it according to each visitor. This is done by first collecting as much data as possible about the visitors and then using that data in order to personalize the communication throughout the user journey. So the key feature of any personalization system is a solid data collection and interpretation feature. 

What types of data should be collected for web personalization?

As long as it’s publically available or if the user gave their consent, almost any type of collectible data goes. Some of the most popular collected data are:

– Technical user details: IP address, name of the company that owns the IP, browser version, type of device, operating system

–¬†Location-based data:¬†user location, current weather at user location

–¬†Behavior:¬†time on site, number of visited pages, scrolling activity, etc.¬†

–¬†Data stored in cookies:¬†if the¬†user is logged in or not, gender, name, previous website activity (such as conversions), etc.¬†

How to create a website personalization experiment?

Creating personalization experiments should stem from the need for increasing website relevance with the aim of improving performance. This is constantly tracking the¬†results of any web personalization experiment is vital. The first thing to do before starting an experiment is to deeply analyze your audience and¬†divide it into different segments that should have their own type of personalization. This way you know exactly who you’re targeting with the experiment and what you’re saying to them.¬†

For example, when you have someone visiting the website for the first time ever, you don’t have as much information about them compared to someone who has been on your site several times already and who’s maybe already a client. So according to the amount of information you have on each segment, you can implement a different type of personalization experiment. 

After you’re done segmenting the audience, it’s time to establish in which way you will customize the experience for each group. Here you have different options from dynamic text & images, to dedicated banners, surveys or even new pages altogether. You just need a tool that can serve the personalization to each segment, and a lot of imagination. 

Website personalization examples

Some of the most common types of personalization experiments are product recommendations according to user behavior.

Another very common type of web personalization is adaptive content, which dynamically changes according to each user. In this example the variable {CITY} would change based on the city of the visitor:

web personalization

Website personalization case study

A short and revealing personalization case study is that of the home & garden e-commerce store Covera that implemented website overlays with adaptive content in one of their campaigns. It was during the winter and some parts of the country were experiencing warmer days (not everywhere though). So using¬†Omniconvert’s¬†segmentation engine, they chose to display the overlays only in the areas of the country that were warm enough for planting new shrubs. Furthermore, they included the local temperature in the overlay’s copy for a complete personalized experience. When the results were in, people that saw the overlay converted with 60% more than people that were not involved in the experiment.¬†Click here¬†for more website personalization case studies.

website personalization experiment

Website personalization tools

Website personalization is quite a fluid concept, so there is a wide range of tools that can help you achieve a degree of personalization. Depending on what you’re looking for you might need a recommendation engine, and¬†A/B testing tool, an¬†online survey tool¬†or some tool that can help you implement website overlayers. A good solution for many personalization needs is Omniconvert, an all-in-one conversion optimization tool that has a wide range of personalization features such as¬†advanced segmentation, A/B testing, dynamic content replacement, website overlays, branching logic surveys and more.¬†Signup here¬†and give it a spin.¬†