 More than just a fancy graphic, the Web analytics funnel can provide you with clear, actionable insights. The big challenge for Web site owners when it comes to improving conversions are high bounce rates, conversion path drop-offs and exits. A funnel report can provide insight into visitor behavior which helps to improve Web site metrics including conversion rates.
There are many funnel reports in various analytics packages, but few have the capabilities of Lyris HQ's Web Analytics Solution. It is one of the favorites of our analytics group at Position² because it has resulted in increased conversion rates of 15-40% for our clients.
In this article we'll discuss Lyris HQ Funnel Report key features, as well as best practices on how we have used it with our clients.
Multi Dimensional Funnel Definition
Most Web analytics tools assume that a visitor’s progression to a goal or conversion page is linear. The visitor starts at the landing page, clicks on Buy Now, fills in the form, completes the purchase and reaches the Thank You page. However, studies have shown that visitor behavior is not linear, but far more complex. Visitors navigate back and forth between different sections such as About Us, Privacy Policy, or move between different product pages before choosing one. This behavior in the extreme is called "pogo-sticking" which has been shown to reduce conversion rates. A multi dimensional funnel analytics tool like Lyris HQ can be configured to take this behavior into account. This leads to more relevant and results-oriented insights. Analytics experts such as Avinash Kaushik of Google have consistently rated this funnel report tool as one of the best.
A key function of the funnel report is to track a visitor’s path from entrance to conversion. It sheds light on the following:
1. Paths that people take towards the goal. This can help eliminate redundant steps. For example, if the report reveals that most visitors tend to read the About Us section, it might be useful to have a summary of the company overview on relevant pages.
2. Exit points in your site. These provide insights into reasons why visitors leave and which sites they go to. Exit points in the funnel are highlighted indicating where visitors drop out in the conversion process.
3. Click Patterns. For example, if you find that visitors click on the Contact Us button across pages, then it may indicate that the relevant information is not easily accessible to the visitor. You can then modify your Web site accordingly.
4. Usability experiments. Another use for funnel reports is in usability experiments. You could create different versions of key pages like the landing page or registration form and see which design has the best visitor progression to deeper stages of the funnel.
Key Features of the Funnel Report
Let’s use examples from a consumer financial tools site to highlight key features of the Lyris HQ Funnel Report such as visitor progression and visitor segmentation.
Visitor Progression
To understand the flow from start to end, you need to map visitor progression across stages. Each stage is defined as a group of inter-connected pages, not just as an individual page. It displays which page groups enable visitors to progress to a deeper level by measuring the progression rate. For example, if a page group is requested in 100 visitor sessions and 50 of those sessions result in a visitor going deeper, the progression rate for that page group is 50%.
In Lyris HQ, this is indicated by different shades of blue (as seen below). The darker the blue, the more influence the page has on driving visitors towards the goal. In the example below, the "Home Page" and the "Sweepstakes Landing Page" are the two most influential pages.
Visitor Segmentation
Segmentation is created by using labels. Using this mechanism, it is possible to study how different groups of visitors react on various page groups. Progression data for each label can then be viewed by clicking on the label name at the bottom of the report.
For example: Stage 1 can be defined as a group of landing pages. You can analyze which landing page had the highest visitor progression for all visitors as well as which page had the best progression for "PPC visitors" vs. "Email campaign visitors."
In the example shown below, "Social media visitors" (segment shown in green) had the maximum progression from the Home Page. (click on image to view full size)
Testing The Funnel
We conducted an experiment for a client who wanted to test the existing form completion (which was all on one page) against a registration process broken out over a couple of pages. The results were striking enough to re-evaluate the registration process. The current "Registration Page" was on a single landing page with eight fields. The modified registration form ("New Register") was split across two pages with four fields on each.
Overall, we found that the progression rate improved from 61% to 91% when we used the two step form. (click on image to view full size)
However, this was not the same for all visitor segments. The progression rate for visitors landing on the home page via organic and social media sources was higher on the New Register two-page form. In contrast, visitors who came in from paid advertising sources continued to show higher progression on the current one-page registration form.
Based on this critical insight, we recommended and helped implement different registration pages based on the visitor's entry page. The result – overall conversion to registration increased by 23%.
Best Practices in Defining the Lyris HQ Funnel
Defining the funnel correctly makes all the difference. Here are some best practices:
1. The first stage of the funnel must be your entry page. This includes the home page.
2. The pages should not be distinguished on the basis of their tracking URLs or parameters since that distinguishes a visitor source and not the page itself.
3. Use labels to create different groups of visitors and analyze visitor progression for each group.
4. Defined page groups should share common attribute. For example: A group of "product pages" or a group of "registration pages." You should also analyze if page modifications have improved the progression rate.
5. Note that if a visitor visits two pages in the same page group, say "Landing Page1" and "Landing Page2", then the progression is assigned to the first page group (the one on the extreme left in our example). Thus interchanging the order of page groups can change the funnel data.
As you can see, analyzing visitor behavior on your site provides powerful insights on how visitors interact with your Web site. It can be a lot of fun as well.
###
Related Resources:
|
