Phone numbers and CRO

I’m always asking does the solution we currently use make it easy for people to transact or can it be improved? Recently I’ve been looking at how we present phone contact information on one of our ecoms. We display a 1800 number in the site header and directly under this a local 02 phone number. I’m assuming to try and lower the cost of providing a freephone 1800 number or for people on mobile phones. It looks messy.

Also we repeat this in our site footer but for some reason exclude the area code for the local number. It’s inconsistent and requires too much thinking for something that should be simple. Our job is to make the process of transacting as simple as possible and remove any barriers to this, regardless how small they make appear.


I think I just described CRO.


How can we improve how we display this contact information? Firstly always establish a baseline of current performance. We’ll use this to compare against our proposed changes, if the changes are an improvement, implement, if not rethink or drop.

I’ve tracked the volume of calls generated from the 02 local number in the header. This accounts for 0.57% of all calls generated, a tiny amount. Now consider that 0.37% of all site visitors will use this information. I question whether we need to display it to all visitors.

Next we track the footer local number to see how many calls this generates. We’ll also need to track the 1800 number located in the site header and in the footer.


Once we have this baseline we’ll setup a sitewide A/B/C test for our proposed changes.

A: Existing.

B: Removing 02 but keeping existing 1800 number.

C: Removing 02 and replacing 1800 number with a new 1300 number.


What I’m interested in testing is the effect of cleaning up this section of the site header to add to cart and order completion rates. We’ll use this, as well as any potential cost savings, when we propose these changes to the business.



I think we’ll see a marginal increase to site conversion rates and a marginal decrease to bounce rates across variants B and C. I also think we’ll see a reduction to call costs using variant C.



What we aren’t tracking are conversions over the phone. While our call tracking system does allow this getting a customer service team to start using the system takes time. For this test we’ll need to work closely with our customer service manager to keep an eye on conversion and call rates, just in case.

Leave a reply:

Your email address will not be published.

Site Footer