We’re slowly building out a brand-new checkout process for DynDNS.com for which we have several main goals: increase conversion rate from each step of the process, encourage more sales using upsell text and other incentives and ultimately make the whole process more user-friendly and efficient.
It’s a pretty simple checkout process and one that an average user will already be familiar with. However, we still see that users encounter some problems and get stuck or leave in the middle!
The main issues we wanted to overcome were:
- Exit rate from checkout pages
- Number of errors on checkout pages
Your unique visits for each page will give you a good sense of how many people are seeing the same page over and over again. If you have a high number of visits (or pageviews) for a checkout step but a low number of unique visits, this means that people are attempting to move to the next step but are encountering some kind of issue that prevents them from moving forward.
In our case, this meant people weren’t entering required information, weren’t typing enough numbers for zip codes, etc.
Track errors using Google Analytics
We agreed we needed to see exactly what kind of validation errors people were seeing, so we set up an event tracker using Google Analytics. The event tracker tells us the actual message that triggers the event like “Missing Zip Code” or “Invalid Phone Number”. We started tracking the errors and saw a big issue right away: there were a ton of people who were incorrectly entering their credit card expiration dates!
Add example syntax
After digging into the issue, I realized that we could really benefit from an example of expiration date syntax. We added MM/YYYY to the right of the field, and this helped a great deal. We also improved the validation to accept other possible date formats (MM/YY, M/YY, etc.) as people entered them to make sure they would still be able to move forward.
This cut total user errors by 80%, but seeing that there were still error trends, we pressed on!
Highlight unusually-required fields
The next biggest trend we saw were people skipping the phone number field altogether. We thought we’d done a good enough job of indicating that this was required, but it appeared that users still didn’t believe it. Phone number is an unusual field to require, but in our case, it helps us to prevent fraud.
I added an extra element to this field: a “Why Required?” tooltip! Though we’re still seeing people ignore it altogether, this has cut down on the number of them significantly.
Check errors by country
After those two main fixes, we continued to see a trend, but not one we understood: there were still a lot of zip codes being entered incorrectly. We have validation for a number of different countries and thought we were doing a pretty good job of catching all the variants – but in fact, we weren’t!
After viewing the Zip Code error by country (in Google Analytics, click on “None” when you’re viewing an event, then choose “Geographic” and the level of detail you want) we found out that the majority of people seeing errors were in the UK.
We revisited our UK postal code validation and found a number of ways to improve it. And voila, significantly fewer zip/postal errors!
Checking your site’s form validation errors using Google Analytics can be an incredibly helpful way for you to improve your site’s usability and keep your customers happy. Dig into it and see where people are encountering road blocks and what you can do to fix them!
Lara Swanson is a web developer for Dyn, an IaaS (Infrastructure-as-a-Service) company that features a full suite of enterprise DNS and email delivery services. Follow her on Twitter: @LaraSwanson and @DynInc.