In terms of metrics people should be looking at bounce and exit rates for individual pages of their websites or apps. This will give them an indication of whether people are finding the information that they want.
In terms of more advanced metrics Page Value is an important one. This is where Google Analtyics uses information from conversion pathways to calculate how valuable (in terms of $) each of your pages are.
Aside from individual metrics, people often forget to look at how users navigate websites. This is obviously key, and gives a much more detailed picture than just focusing on individual pages.
We had a user visit the site and ask a question which I’ll post below (it’s a long one). Do you have any advice for this user?
We have a corporate website built on Wordpress and a web based product offering that our users log in to which is built on a different tech stack and hosted separately but all on the same domain.
Our marketing team decided to take the LOGIN/SIGN-IN link off the corporate pages on the home page as well as all other pages stating that the corporate website caters to corporate clients and they would get confused seeing a sign-in link on the website.
Our user base is obviously larger than our client base and hence my argument that we need to have a Sign-in link on the homepage and on all pages on the corporate website which is also what our users are told to go to in our marketing campaigns. Currently the Sign-in link is only on one page on the corporate website which talks about the product we offer and no where else. Getting to that page takes some scrolling and 3 clicks after someone lands on the home page.
But our marketing team is asking me for any metrics or data that says that users are confused doing the above actions to get to the Sign-in link. They believe that our users who need to reach the login page should bookmark the login URL if they need to get to it faster - but the product is offered to everyone on the web so I can’t train them all so easily to do that.
How should I give them metrics they ask for? From where I stand, I feel this is a no-brainer that we need an always visible Sign-In link at the top of the home page but am I asking for too much?
From what you’ve said, I agree, it does sound like a no-brainer. However, I’d advise an analytics first approach. Using analytics here will give you a much stronger argument, as it’s harder to argue with data than it is to argue opinions.
Firstly I’d look at how many users are landing on your sign-in page; these will be the ones who have bookmarked it. You can then work out the percentage of visits that come directly to this page (i.e. what percentage are using bookmarks to find it).
Next I’d look the navigation paths to that page to find out whether users are entering on the homepage then navigating to the sign-in page.
Finally, I’d recommend (if possible) running an A/B test to see how user behaviour differs on the design with a sign-in link on every page, vs. one with it only on corporate pages.
…also, from a UX point of view, you could make it clearer that the sign-in link is only for existing customers. This could be as simple as labelling the link ‘Existing customer sign-in’ or similar. Begin with the data though to find out how people are currently using it.
Great! thanks for the solid advice.
@resources, I hope this helps your situation going forward. 
I have not read the book, so please forgive me if I’m asking in ignorance.
My impression is many use GA to research the effectiveness of various links to the site. That is, the focus is on the external. So it is interesting to know that GA can be used to research internal as well.
A problem I have is I’m usually “too close” to my own stuffs. I have my own use habits and I know if I do X, Y will happen, and so much of what I do is rote that I don’t even think about it on a conscious level.
One thing about GA that I have been interested in but have yet to get into is setting up AB testing for things like “which is getting used, the nav menu or the content link”
In your opinion, would setting this up have a steep learning curve?
This is the exact kind of thing that people should be using their analytics data for. Internal navigation on a website can be seen in a couple of different ways in Google Analytics. There are flow reports, which show a range of pages and how users navigate between these. There are also navigation reports for individual pages. These reports are available in GA by default, and aren’t too difficult to use.
As well as these reports you might also want to consider using heat mapping tools to see where people are clicking on your pages. I often see through heat mapping software that more people are engaging with footers than you might expect, particularly on mobile.
Here’s a good introduction to navigation reports in GA. Also, I recommend reading the book for some examples of how I’ve used navigational reports to gain insight on previous projects )
Thank you @lukehay for joining us today and teaching us a more about Google Analytics, and how we can dig for treasure with data.
What’s the next big project you’ll be working on? Will you be following up with another book?
I enjoyed writing the book, but I’ve got no plans to write another one for a while. I’ll keep writing blog posts, but it’ll be a while before I write anything as long as that again.
The projects I’ve got lined up are helping to organise UX Camp Brighton, running my own > analytics and UX training workshops and continuing to help people understand how users are interacting with their websites.
Excellent, sounds like you’ll be keeping busy. Where can users find your blog posts?
A few places. I write guest posts for sites like UX Booth and Conversion XL, I also write for Fresh Egg and finally my own blog, which I’ll spend more time on now that the book is finished!
I just downloaded the book.
Do you have written somewhere, specific case studies of this stuff? A real world, practical example from top to bottom analyzing data and making certain changes based on the data and measuring results?
It’s one thing to generalize “this page isn’t performing well” from GA, but it’s hard to pull out the “why” from GA alone. If one sells two products, a $20 one and a $2000 one, probably the $20 will perform better simply because it’s cheaper. Just because the $2000 doesn’t “seem” to perform as well, doesn’t mean there is some kind of SEO or other problem that GA will reveal.
I would guess there is still a lot of speculation regarding interpreting what analytics is saying! So I’m most interested in the types of data GA can reveal that doesn’t involve speculation. Like finding actual problems.
There are quite a few case studies in the book. I wanted to make it as practical as possible. It’s based on my personal experience rather than hypothetical examples.
You make a great point about pulling out the why from GA though. The simple answer is, in most cases you can’t use GA on it’s own to fix underperforming pages. In fact the main purpose of the book is to encourage people to not just rely on analytics data but to use it alongside other UX methods. The analytics data should be use to find where problems might be, other methods should be used to confirm whether these are actual problems, and also to begin to understand how you might go about fixing them.
While analytics data can be really useful, I’d never recommend relying on that alone. A qualitative approach is needed alongside the quantitative one.
Thanks @angelamolina for pitching in this question for me and @lukehay for your tips. I think those are valuable inputs towards my cause and I will find ways to get through to the branding team with my research and data.
Looking forward to this! Signed up just to check it out
This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.