Google Conversion Room Asia-Pacific Blog Tips on tracking and improving conversions online

Shoes of Prey - Using Custom Reports to identify influential pages

Tuesday, January 25, 2011 | 10:13 AM

Labels: , , ,

Michael Fox is a co-founder of Shoes of Prey, an online retail store that allows you to design women's shoes online, which are then handmade and shipped to you. Michael blogs about running Shoes of Prey at the business blog www.22michaels.com. - Ed.

Like most online retailers, we’re always looking for ways to turn our website visitors into customers.  We use Google Analytics to monitor the performance of our website and our online marketing and to understand how to make our website even more appealing to potential customers.

As part of our ongoing efforts to improve our website, we recently asked ourselves: which pages on our site are most influential to our customers’ purchase decisions? For example, if someone visits our testimonials page, does that make them more likely to make a purchase? How about watching the videos on our leather page, visiting our gallery, or arriving on our small shoes or wedding shoes landing pages?  We knew the answers to these questions would help us maximise the number of purchases from our website.

Custom reports
To start answering these questions, we created a Custom Report in Google Analytics to review the “Highest Converting Pages” on our website.  You can make a copy of our custom report setup here and use it with your own site data.






This report shows us three key data points:
  1. Pages that people visited (Page column)
  2. The number of views each page received (Pageviews column)
  3. How many sales (ie goal completions) were generated in that same visit session after visiting one or more of those pages (Total Goal Completions column)

Analysing the data
We then divided Total Goal Completions by Pageviews (on a separate spreadsheet) to discover what percentage of people who visit a particular page end up making a purchase. We assumed that the higher the percentage, the more valuable that page for turning visitors into customers. We could then promote the content on those pages more across our website.



We kept in mind a few circumstances that could impact this data. For example, people visiting our shopping cart or logging into our members area will have a high conversion rate, so we filter those pages out. Customers looking to purchase may be more likely to visit our page about our returns policy. We also kept in mind that we were looking at the same visit session only. Often, customers make multiple research visits before they finally make a purchase.  Therefore, we were conscious that there may be other pages that motivate purchases, but are not reflected in our custom report because they were viewed in another session.

After reviewing the data, we determined that our most influential pages were:
  1. Leathers page
  2. Returns policy page
  3. Tab2 - the delivery tab on our product page
  4. Customer testimonials page

Taking action from our insights
Knowing our most influential pages has helped us maximise the number of purchases we generate through our website. Once we reviewed the data, we took the following next steps:
  • We’re in the process of redesigning our shopping cart, and we’ll look at ways to incorporate this information within each step to reduce drop off and increase conversion rates. For example, we’ll provide information about our returns policy and deliveries within the shopping cart pages itself. This will reduce the need to click away to our FAQ page.
  • We’ll also consider making some of this information, especially our customer testimonials, more readily accessible on our home page and custom shoe designer.
  • We’ll include some of this information on key search engine marketing landing pages such as our gift certificate page.  For example, we’ll incorporate customer testimonials in order to assure those purchasing our gift certificates that they are providing a gift that will be appreciated.
  • This will be good information to incorporate into discussions on our Facebook fan page and in videos on our YouTube channel.

We’re excited by the range of insights that Custom Reports provides us.  It allows us to slice and dice our data within a single report rather than having to combine separate reports within a desktop spreadsheet.  We would love to hear about the custom reports that you find useful for your online business.

5 comments:

Ophir Prusak said...

I totally agree that looking at the correlation between pages viewed and sales can reveal insights, but as you mentioned, it's VERY important to make sure you're looking at the cause and not the effect.

This is covered in detail here:
http://online-behavior.com/testing/cause-effect-and-split-testing-1072

Unknown said...

First, thank you for posting this Blog entry. This is a very useful study. However, I have to disagree with one of the statements made, but not with the spirit of the analysis.

"We then divided Total Goal Completions by Pageviews (on a separate spreadsheet) to discover what percentage of people who visit a particular page end up making a purchase".

When the Goals Completed is divided by number of Page Views for a given period, the answer does not indicate the number of "people" who have converted to a purchase (purchase being defined as a goal completion). Instead, it demonstrates how many average Page Views it takes to complete a purchase on that goal funnel. Page Views are not an indicator of people (which equates to Visitor or Unique Visitor), since the number of Page Views that are possible per Visitor varies per person. This can seem like a technicality on the surface, but since Google Analytics measures Visitors by counting the number of __utma cookies that exist on each physical computer, and each user can both delete cookies, and have multiple __utma cookies on their systems at any given time (one for each browser used to access a Google Analytics tagged website), the actual count of users is variable at any given time. Page Views becomes the only reliable statistic rather than user counts since it truly represents a view, regardless of user or person associated, and is not detracted by the number of cookies on a system -- the more the merrier.
The measure you've outlined very useful, but it represents views of the particular page that is being measured by the goal attributed to it, not the number of visitors who made purchases.

Thank you, again, for posting this article. It was very helpful, and I appreciate the opportunity to add my comment.

Chris

MM2 Analytics

Michael Fox said...

Chris - thanks for the feedback. What you say is very valid in terms of the difference between pageviews, visits, and visitors. We were using pageviews as a proxy for visitors/visits leading to sales and didn't mean to use "people" literally in this context.

Rakesh M said...

Hello Michael,

We have automated the reports that you have built and equipped it with some additional features using Tatvic Google Analytics excel addin. We name it "Page Importance Level Scorecard" and it can be downloaded from Tatvic Blog.

Arnold McLean said...

Great reading your blogg post