A Typical Conversion Project

Installing Conversion Tracking Software

In order to measure improvements, we will first need to install some form of conversion tracking.
This can be a paid solution, which will need to be costed separately, or we can use ‘Google Analytics’, which is free.

Tracking User Behaviour

Before any changes are considered, we need to analyse the behaviour of your site visitors to see if we can identify any obvious ’sticking points’.

We do this by installing software that tracks and records how your visitors interact.

We will be recording and analysing:  

• where they click their mouse
• how they move the mouse around the site.
• the areas of the site that they spend most time reading (measured by scroll patterns)

This will require a couple of lines of JavaScript adding to the home page and each page of the form. There should be no impact on site performance.

Depending on the number of site visitors, this will usually provide useful results within 4- 5 days.

Nevertheless, the testing will be run continuously for a period of a month, and any secondary conclusions used to refine the other stages.

Visitor Profiling

We work with you to identify ‘typical’ client personae. 

This is a process of creating one or more imaginary characters that encapsulate the  desires, hopes, fears, skills, limitations and preferences of you potential clients. We want to identify what is likely to motivate them to act and what obstacles are likely to prevent them from acting.

All proposed changes will be with reference to these visitor profiles.

Research

Other web sites are targeting your customers and some of them have
already completed conversion improvement programmes and will have
devised effective ways of improving the ease of use and rate of
conversion.
We will search out and analyse any sites, within and outwith your
market sector, and identify any ‘conversion factors’ that may help to
improve conversion for your site.

Real-Time Usability Analysis

We will hold one-on-one usability-study sessions with volunteers chosen to match your user profile. This will allow us to identify further usability issues that may need to be addressed.

Usability Report

Based on the data received from the previous stages, and incorporating web site usability best practices we will prepare and submit an interim usability report for discussion.

This will detail site elements that appear to be hindering the conversion process and also any additions that may improve the conversion process.

Consultation Session

Once all parties have had an opportunity to read the usability report, we will hold a consultation session in order to plan the split run testing stage.

This works best as a ‘brainstorming’ session, with all concerned parties invited to contribute ideas that may address the issues raised in the usability report.

By the end of this meeting, we should have agreed on:

• A primary list of site modifications that we wish to test out immediately.
• A list of secondary modifications that we can consider for testing later.

Split Run Testing

Once we have agreed the primary modifications for testing, we will set up the campaign on our Multi-Variate Split Run testing software, and work with your web design team to insert the correct html on your live site. This is normally a fairly straightforward process with only minor changes to the html code required.

The testing process will provide useful results to analyse within the first few days, but we recommend leaving it running until such time as the results are statistically significant.

This will depend on how many factors are being tested, but for a site with reasonable traffic, should take less than a month.

Results Analysis

Once we have achieved acceptable results from the split-run test, we have the option of ending the test and going live or starting a new test.
Any new test would incorporate new ideas that have arisen during the test process and any of the previously agreed secondary modifications that are still relevant.

This process can repeat as often as required within one month of the split run test starting.