Segment visitors by intention with Google Analytics

As I mentioned before, understanding what visitors want and giving it to them is the key to a successful website. That is the big picture.

Now let me tell you how to actually measure this. My tool of choice for this is Google Analytics.

With Google’s Adwords Conversion Tracking you can define goal pages and track conversions that happen once visitors land on those pages. For example: thank you pages for signing up, downloading a white paper or for purchasing a camera.

Many e-commerce websites have a multi-step check out process. Once you hit the “buy” button you are taken to a page where you can select the quantity of the selected product and other variables. After this you are taken to a page where you input your shipping information. Later to another page for you to input your billing information. Followed by a confirmation page and finally to the thank you page. This is commonly known as the “conversion funnel”.

You can use funnels to identify and reduce drop-out rates throughout the conversion process. Google analytics provides tools to create such funnels and reports to measure them.

The main problem is that most people optimize their conversion process, but don’t measure and optimize their persuasion or pre-selling process as well.

Once a visitor clicks on the “buy” button, he or she is already set on buying the product, but the path to conversion starts way before that, it begins with the persuasion process. I explained that process on this post.

In short, visitors come to your site with a specific mindset (expecting something in particular and the keywords they type are the best clue to what that is). It is important they land in the right pages and that those pages induce them to move to the next step in the persuasion process.

Now let’s see how we can use Google Analytics to segment your visitors based on what they want.

In addition to segmenting users along pre-defined segments such as geographic region and language preference, Google Analytics allows you to define custom segments and analyze the behavior of each segment. For example, you might ask visitors to select their job category (such as Engineering, Marketing, motorcycle stunt riding, etc) from a form. You could then analyze browsing and buying behavior based upon the selected job categories.

What we are going to do is create three custom segments: On for visitors seeking primarily information or products with no specific brand, another for visitors looking for a specific brand, and another for visitors that actually want to take action (buy, download, subscribe, etc.). Those visitors are expected to land on relevant web pages, if we set our PPC or SEO campaign right. We can then label those pages with their segment (generic, brand, action).

To do this, insert the following code to each of the pages.

For pages visited by people looking for information and not a specific brand, i.e.: people searching for “laptops”:

<script type=”text/javascript”>__utmSetVar(’Generic’);</script>

For pages visited by people looking for a specific brand and model, i.e.: people searching for “dell latitude”:

<script type=”text/javascript”>__utmSetVar(’Brand’);</script>

For pages visited by people who are looking to take action, i.e: people typing “cheap dell latitude d420”:

<script type=”text/javascript”>__utmSetVar(’Action’);</script>

These codes need to be set below the Google Analytics tracking code.

You can view conversion behavior for each of your custom segments in the Marketing Optimization > Visitor Segment Performance > User-Defined report. Also, many other reports allow you to cross-analyze data according to visitor segments. To cross-analyze, click on the Analysis Options button at the far left of any entry in most reports. Select: User-Defined, to view your custom segments.

Cross-analyzing, using this segmentation, can provide actionable data that can be used to improve the persuasion process and lead more visitors to the conversion funnel.

1 reply

Trackbacks & Pingbacks

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *