Creating an optimisation department – Conversion Conference 2013
Conversion Optimisation

Creating an optimisation department – Conversion Conference 2013

23rd October 2013

It’s been a good day so far at Conversion Conference, next up we have a session on creating an optimisation department by Oliver Palmer and Pete Taylor.

First up we have Oliver Palmer who is talking about getting buy-in for testing in a large organisation.

Often, it can be really hard to influence change in big organisations who may not be listening to you. Implementing a CRO strategy is in a small business is a lot easier, there aren’t many stakeholders, everyone is more united and you can quickly implement tests.

Doing the same in a large corporation is more difficult. Different groups have different priorities and they have different incentives and even if you’re well versed in collecting data, making sure your voice is heard can be hard. Oliver talks about a test he ran at a mobile company where his test was successful but where one stakeholder didn’t like it because he was aligned differently to the goals of the test.

Key point from Oliver is that you need some quick wins, to test some obvious pain points and to use

Quantify your value

Once you’ve got a few quick wins, you need to quantify your value to the business and talk the language of the business. If you want more resources to test, then you need to do this. Avoid using conversion-speak, speak the language of their business such as orders or customers or engagement. Ultimately though, turn it into cash.

Create evangelists

Try to create internal evangelists who you can help with testing. Speak to them and find their problems, work with them, run a test and solve the problems.  Quantify the gains that you make and allow that person to show their boss what you both did, so you’re making them look good and they can really help you get stuff done.

Share results

Create a distribution list of the people who need to know about the results that you run – even if you don’t get big wins. Hold regular meetings that involve different people across the business and share what you’re working on. Show them that you’re available to them as well.

Also send out “which test won” style emails that show very clearly which tests have won. People make the wrong assumptions and are often wrong with their gut feeling, so show this to people and show them the ins and outs of doing conversion optimisation.

Second up in this session we have Pete Taylor from The Student Room and he will be talking about how they implemented CRO within the organisation over the last 12 months.

Going back to 2008, it was a much smaller business but they recognised that CRO was something that they needed to work on, but there were pitfalls along the way.

Where they went wrong

They engaged a consultant who gave them some poor advice. They discovered and one size fits all wasn’t the right approach to CRO and their technology was not up to scratch. They also learned that letting tests run their course was important and they shouldn’t be ending them early. As a result, CRO was left to one side and people lost faith in it.

When things changed

At the end of 2012, one of their sites had been hit by a Panda update, ambitious growth targets were set which led to them revisiting CRO again.

Pete had to start looking at how much to budget for this and had to look at getting an external consultant. From experience, he also knew that the technology had to be right this time too. Having development time was also important so that tests could be run quickly and easily.

Pete emphasises the important of the right consultant who understands the business.

Stakeholder buy-in

CRO is good in that you can show when extra money has been made, however not all tests will yield the financial result that you’re expecting. If management see the process as a black box, it can be hard for them to understand and give you buy-in.


Look at your analytics data, getting survey results, run user testing if you can. All of these you can use to build a hypothesis for the testing which management can feed into, this lets them see part of the process and they can put their input on the test too.


Pick your tests and get them started. Pete prefers to run tests in certain orders, he picks tests where he can run multiple tests at the same time for different types of users. You then need to be able to change a test quickly if necessary.


Perhaps the most important of all because not all tests are going to be winners, but even the losers can give you data that you will learn from. The team will dig into exactly what worked, the picture it has painted.


Keep testing, you’ll never get to 100% perfect so you need to keep testing and trying to beat your previous tests.

Marked by Teachers saw a 78% uplift in conversion rate as a result of using this process. At this point it becomes a lot easier to get senior management buy-in to keep things going and test more. But the hard work continues and this success and help to push through tests across their other properties like The Student Room.

Where next?

Pete wants to remove silos which can cause problems in going through the process. He wants to have a dedicated role and team within the company which will include dedicated development time, on-going analytics health checks and to always be testing. It is important to have dedicated development time and making sure that accounts are configured correctly such as Google Analytics being tweaked to your needs.

Written By
Paddy Moogan is co-founder of Aira, a digital marketing and web development agency based in the UK. He is also the author of “The Link Building Book”.
  • This field is for validation purposes and should be left unchanged.