The four Pitfalls of CRO
Dig a little deeper
When people first get to grips with analytics, they usually look at the top numbers quickly and make assumptions without digging into the data to find out what the real issue may or may not be.
At it’s very best, headline figures are just that: A headline. Much like the title of a new chapter in a book, it gives you an overview very quickly, but it’s only when you get deeper and more involved into the narrative you realise that the title has little to do with the story other than as an intriguing placeholder or generic overview.
We’ll use a common example:
Checking your analytics you see that mobile traffic is higher than both desktop & tablet, sitting at 57% of your total incoming traffic with desktop only taking up 18% of the traffic. This is the exact point that someone somewhere utters the words “let’s put all our resources into driving mobile traffic to the website”
This is the point where you question, question and question again.
The first thing to look at is the Behaviour of mobile users vs desktop vs tablet users. Look closely at the average session duration, the pages per session, the page depth and to some extent the bounce rate (if you’re working with a site that has a required goal it might be worth looking at the conversion rate of said goal).
Continuing the example above, the average mobile session is 09 seconds compared to 3 minutes 28 seconds on desktop. Mobile users are viewing 1.14 pages per session compared to desktop at 3.12 pages per session. If you include the bounce rate, mobile is 93.16% and desktop sitting around 26%. You don’t have to be a mathematician to see that throwing huge amounts of revenue at mobile is going to bring you a lower ROI than desktop in its present form.
My advice for this situation: focus your marketing on desktop whilst you find out what the problem is with mobile, that brings us on to my next point.
Don’t put design in the corner
We use the term design loosely; this can cover UX, front end and artwork, but the premise is still the same.
Again, most probably the same person as before who told you to plough all your resources into generating mobile traffic will pipe up and say — bloody design! Fix the design. Now I’m not saying that this person is wrong I’m just saying they could be wrong; some examples of digging deeper would be to look at the following:
Look at heat / click maps to see what, if anything is awry. Find out if any of the mobile browsers are under performing or have a higher percentage of negative stats as this could be your problem.
But more importantly, which is nearly always missed, take a look at your marketing. Your website analytics should be linked up to marketing efforts to map any changes in acquisition and behaviour. If the marketing coms do not match the message on your website then people will land and leave.
Check yourself before you wreck your stats
Quite often we’re required to look at a company’s analytics so we can make recommendations. The first thing we do is block our IP address, which is right around the same time we realise that no other IP addresses have been blocked. In a nutshell, this means that all the updates to the site, all the visiting and checking by your team (God help you if you have a call centre) is included in your statistics, so you don’t get perfect view of what’s going on when your customers visit!
Not only do you have to block yourself, but you should also block any analytics software that may be pinging your site, as well as countries you don’t market yourself to.
Points of view
When you block traffic to your site, you need to do so in a separate view. All too often, people will make changes to the master file which contains all the data.
You should create a different view for different types of data that you are trying to capture. For example, you may have a website that caters to different types of users. If you’re an education establishment, you may want to differentiate between internal traffic and external traffic that comes to the website.
The benefit of this is that all your data is segregated, so you can gain a deeper understanding of your audience and when you need to you can always go back to your master file to get all of the sites data.
This is the simple part and can be summed up as: question everything, assume nothing and don’t forget to test your hypothesise. Rarely in my career have I seen assumptions turn into facts. Fact.