member of

Behavioral targeting in 3 (not so) simple steps

So you manage a website and thought Web Analytics was advanced? Get ready to start working on the next level of analysis, this time in Real Time: Based on the behavior of the visitor you’re going to be instantly serving them the right, personalised content. At least...That’s the theory.

Behavioral targeting is not new. For a long time it used to be as simple as tracking the websites people visit and then serving them ads based on that. That’s Ad targeting. This type of targeting has now been extended to also consider all other type of information you leave behind surfing the web and even information that is collected about you or your neighborhood offline. Specialist companies that sell this data are Nielsen and Comscore but also Google and Yahoo are in this business. Sites can become ever more personal this way. This type of targeting is dubbed network behavioral targeting and leaves many worried about their privacy.

No, the type of targeting we’re focusing on here is On-site behavioral targeting. It uses information such as how visitors got to the site, the number and the kind of pages they view, if they are new to the site or have visited before, and the countries they are from. This is all data which is available in most analytics package.

Getting started

We will be focusing on using free tools. All you need is time, dedication and brainpower.

1. Research: Get to know your visitors and put them in (initial) segments

What you want to arrive at is a small number of segments of users that you want to serve different pages to. Start by defining a few important segements, you can always add more later. You need to be able to trace the needs or attitudes of visitors in each segment to data that is available to you. Creating segments can, and should be done from a number of different angles:
By analysing how visitors currently use your site. For this Google analytics is a free an very capable tool. Tip: The search terms used to find you can be very insightful.
By asking your users about their wants and needs, directly via an email email survey or by putting up a survey on your site. This helps to deepen the understanding of the segments you are trying to define. 4Q Survey is a free and effective tool for this. If possible you can also interview your visitors in person.

If you arrive at insights that challenge your current content, functionality, product offering or even your business model do some thinking and work in this area before continuing.

2. Design & Develop: Create different designs for each segment

Now that you created the right segments in Google analytics you can use them in the free BTBuckets tool. Here you can define changes to the page or actions that should occur once a visitor is matched to a segment. Segments are called buckets in this tool. You can choose to offer the different segments different text or images, links, pop-ups, banners, redirect them to a different page etc. Deciding on the right actions for the right segments is where a lot of thinking happens. You suspect a segment will respond better to a specific design but you should consider your interventions as hypothesis until they have been evaluated in step 3.

3: Optimize: Evaluate how you are doing, make changes and iterate.

Your analytics package can track the performance of different segments for you. This way you can compare the performance of these segments before and after any changes (assuming your goals can be measured). But there are quicker ways to optimize: You can make 2 different designs for a single segment and use Google’s website optimizer’s A/B testing or even their multivariate testing in conjunction with BTBuckets. You can see which design works best and discard the others. It is also important to keep evaluating whether the segments you created in step one are useful enough or if you should change the segments to better fit your and your visitor’s goals.

Lineup of Tools

  • BTBuckets is free up to 5 million views/month and has a nice integration with the free Google Analytics and Google Optimizer tools
  • Personyze offers more features, making it easier to customize the pages in step 2 and optimize actions in step 3. It is free to try for 30 days and plans start at $59 per month
  • Sitespect is an interesting tool because it does not require any change (coding) to the site content, everything is handled by Sitespect.
  • Three well-known, more expensive packages are WebTrends, Omniture/Adobe test & target and Coremetrics
  • A big overview of tools

Need some help?

Please get in touch to find out what we can do to help you get on your way