Personalization - Response Based Segmentation Strategy

Personalization and Segmentation Strategy

You only have one chance to create your first impression. Selecting the main message for the homepage of a website with two very unique target audiences can be very complicated. You can select one of the segments and create a customized experience for that audience or you can negotiate an image that could work with both segments. Either way - you are probably not creating the best experience for each individual segment.

How to design homepages for two distinct audiences?

I saw this happen recently with a client in the health industry, where site visitors were both future dads/moms and Physicians. The website is about a biotech product for expecting parents, so they have a strong emotional client pitch. But at the same time, they must provide a serious technical message about the product for Physicians.

Audience A

Prospects
80% of site visitors
Audience B

Doctors
20% of site visitors

Currently the site targets the actual clients on their homepage (with a more emotional message based on tranquility), but they also included a large call to action for doctors to reach a technical landing page. They wanted to know if we could automatically "detect" these audiences when they reach the site and create unique messages on the homepage. With specific referring URLs (such as technical medical websites) and some organic keywords we were able to deduce the audience and create the experience. However, this traffic was already usually directed to specific landing pages and most of their traffic was brought by their brand name for both audiences.

Ask your users who they are

We then discussed a behavior we saw on another Brazilian health site (bulas.med.br) where you are able to continue your navigation to specific parts of the site only after you select what type of visitor you are using a simple gateway pop-up.

Segmentation Using Popup

By offering a choice of audience segments that a user can define him/herself during the first page view of that visit will let you create a personalized experience. After the user selects their segment - you can already act on this information.

Personalization and advanced metrics

Easy - yes! Intrusive, very much! But what a lode for personalization and audience analysis. Not only can you now offer the appropriate messages right on the homepage, you can also analyze referrers, bounce rates and funnels per audience. Here is an example on how you can filter any report with these clusters using Google Analytics traffic sources segmented report:

Google Analytics Segmentation

If you interrupt, leave the door open

Besides the obvious consequence of some users simply leaving the website because the popup is interrupting the visitor's expectations, here are some other considerations for this type of implementation:

  • Some visitors may not belong to any of the pre-selected segments and could have their experience influenced by an incorrect classification from their side;
  • Some visitors may also be reluctant (or even upset) to give out this information and select an incorrect answer on purpose;

Simply adding a "Don't Know / Just browsing" option may help solve some of these challenges - even if you don't cluster every visitor.

Test the response-based segmentation strategy

By knowing exactly the visitor segment you can create a more engaging user experience. When segment expectations vary - creating an extra step in the user navigation may offer the better experience.

If you are a content website or offering information people are searching for - these users may be willing to give this information. However, if your site is product specific or strongly associated to a brand, some clients may bring certain emotions of uncertainty - which could ultimately hinder any benefit of a customized segment experience.

If your site does really benefit from unique messages, you should test the response-based segmentation strategy. Ask a small percentage of your visitors their segment and then check if bounce rates and feedback forms are affected for these users. Also study the analytical data for each segment for previously unnoticed segment differences.

Online Behavior © 2012