How I Use Excel to Reduce Page Bounces and Improve My Conversions
In my role as CTO of NEXT Analytics, I see a lot of dashboards going through my inbox. Over the years, I’ve seen literally thousands of dashboards, all built by experts from all over the world. I have a confession to make :), I don’t actually use any dashboards for my own business.
Instead, I pull raw data into Excel and I visually scan it. Sometimes I will use conditional formatting to draw my eye to the outliers, sometimes a chart. I don’t use dashboards, I am a bad boy. I find visualizations to be tedious and rarely useful. I spend more time developing the visual and I still don’t end up with any actionable information.
Dashboards aren’t useful to me for most things that I want to know about, specially if I am planning what I am going to work on in the next hour of my workday. I personally prefer ad hoc analysis. My belief is that if I spend one hour every week querying, reviewing, and fixing bounce problems in particular, I will have a high impact on my business goals.
To get the maximum effect for the time I invest, I focus on High Bounce count because I know that at least some visitiors could be converted if I did a better job of communicating with them.
The reasons why they might convert depends on how I’ve segmented the data. It could be the page, the source, the location, or even something technical like html or how the site is hosted.
Since Bounce can be caused by many dimensions, it’s tedious to interactively go through each one. To save time, I developed a single spreadsheet that gets all possible nuances of contributors to Bounce that I can think of.
When high volume page has a high bounce rate, then often the fix is improving the content on the page, making it more relevant. That takes time and effort, so I want to focus my time and effort on the pages with the biggest impact.
- High Bounce Goal funnels amd Event Pages
- High bounce Landing Pages
- High bounce Page Titles – multiple pages using same title, review the content of each
For other dimensions, there are sometimes other issues that require more specialized fixes…
- High bounce traffic sources – review the links for content, images, and also if the source is acting in-appropriately.
- High bounce cities – are there locale issues in my content? Should I exclude this area from my paid advertising?
- High bounce Keywords –If I am getting a high incidence of a high bounce keyword, I should redirect to more specific content
- High bounce Campaigns – review adwords campaign, but also I might also review the use of utm_ codes
- High bounce Ad Content – redirect to specific content, better images, better wording, locale issues
- High bounce Browser, OS, Mobile Devices, Site Speed – review content including images but also html and hosting
Each week, I review all aspects. When I see a problem, I make changes to the site(s). Gradually, the problem pages leave the report and new ones appear, hopefully with lesser number of bounces.
- Better wording and images on the links that drive traffic
- Better wording and images on the pages which have high bounce.
- Sometimes it’s a site problem, either speed, reliability, or how it’s implemented.
I focus my efforts in these areas and the effect is very measurable, very beneficial. Because I started out by looking at high volume pages, the improvements I make are very leveraged. I think it’s a good use of my scarce time.
As an analyst, this is the kind of work I want to be involved in. It clearly adds value and I can prove it by watching the change in my goal and ecommerce conversions.
To speed you along, download this free Excel worksheet (below). It automatically fetches high traffic pages and sorts them by a high number of bounces. It segments Bounce by about twenty different dimensions, each one providing a perspective on the cause of Bounces in your web site.
Use this spreadsheet to analyze the Bounces on your WWW Site. Because it favours high volume, any improvements you make will affect the a large number of visitors. For example:
Start with analyzing possible causes of bounce and taking action.
- create or improve content for traffic sources that are sending qualified visitors who bounce
- improve the content on pages which more generally have a significant number of bounces
- remove traffic sources, especially if they cost time or money, which are generating a high number of bounces
Not Rocket Science!
This is pretty simple data that can have a big impact.
It’s a one click data refresh and takes less than a minute to run on most sites. That’s quick and convenient by anyone’s measure.
Download the Spreadsheet:
Use NEXT Analytics Excel Addin to See Your Numbers
If you don’t have NEXT Analytics software on your computer, get a free trial copy at our corporate web site:
In General, Watch for these…
- Clean up, standardize, your marketing, e.g. establish a messaging platform, choose a tag line, lock down your branding, and stick to it until you see a reason to change.
- If they are arriving from a paid-referral, then you can save money or get better results if you remove those sets which are sending traffic that’s not converting or bouncing.
- Improve the clarity of your banners, ads, and other promotional messaging.
- Refine your messaging so people know in advance what they’re going to see if they click.
- If a valid prospect didn’t find what they wanted in the duration they were on the page, you want to make the pages more informative, more persuasive or better at sending people to pages which did have what they want e.g. internal search.
- A valid prospect didn’t see what they wanted on any visible portion of the screen.
- A valid prospect didn’t choose to navigate.
- A valid prospect didn’t recognize the call to action.
- Were there problems with the coding of the site?
- Is there a problem with the web site hosting
Beware Errant Bounce Values
Yehoshua Coren has a great blog article that helps to understand the phenomena of Bounce. http://www.analytics-ninja.com/blog/2012/06/google-analytics-bounce-rate-demystified.html. In that article, Yehoshua mentions that Justin Cutroni also has a great article. http://cutroni.com/blog/2012/02/29/understanding-google-analytics-time-calculations/. Both authors explain that bounce involves “Hits” which are the accumulated values for events, ecommerce items, transactions, social tracking, and user defined actions.
The gist of what Yehoshua discusses is that a high bounce is often caused by the phenomena of web developers who are trigger Google Analytics “hits” that don’t indicate bona fide actions for the purpose of business analysis. The extra “hits” affect the value of the bounce metric. In short, extraneous hits can drive down the value of the bounces metric. Yeshoshua’s point on this matter was that web site programming often contributes to high bounce rate but since they don’t indicate user engagement they can be misleading. He also makes the point that some pages are conducive to single visits, i.e. bounces, and that’s just fine.
Yehoshua also pointed to an article by Eivind Savio who has proven the benefit of not relying so much on Bounce, but on http://www.savio.no/blogg/a/114/tracking-content-scrollers-scanners-og-readers-in-google-analytics