Earlier this year I touched on how to get started making your website work better. Conversion Optimisation is a continuous cycle, you make small improvements continually and the result is continual growth. Anything you do with consistency, you should back up with a framework. I’ve gone and found the 4 leading frameworks and put them all in one nice little spot for you.
It might be best to go grab a coffee before we start looking at methodical approaches to planning, implementing, executing & reviewing your tests. Fresh air helps too. Or you can jump to my sum up of these ideas here.
Now – On with the 4 Frameworks!
This is Widerfunnel’s method – it’s the next step of their Kaizen Plan (essentially an thourough business analysis). They prioritize on framework of PIE – Potential, Importance, and Ease.
1) Landing Page Analyiss
Here WiderFunnel identifies areas that fall short, uncover conversion gaps and gain overall insights.
2) Hypothesis Creation
Using the information from the previous step hypothesis for direction are created – these hypothesis must be able to be tested, measured & proved effective. These are based on what is causing the conversion block and what the best approach to navigate that blockage.
3) Experiment Map
This is a document that lays out the specifics of the tests. It includes objectives, structure, estimates, number of test variations, wireframes, timing & notes.
4) Graphic Design & Copywriting
Test pages are created by designers, content is done by the copywriters all based on the Experiment Map.
5) Technical Installation
The test variations are implemented, integrating the sites themes to appear less shocking. Tracking codes are also key in this phase.
6) Experiment Launch & Results Tracking
Test time. Results are monitored on a daily basis, the data is used for the best tactical recommendations for maxing conversions.
7) Results & Analysis
Once enough data is collected for statistical significance. Hypotheses are verified and the valuable input is used to lock in a new web page. Then the process is repeated with the new page. It’s important to note that each time you do the process you’re not starting from square one, but gradually building your knowledge base of what works and what doesn’t.
Conversion Science has labelled their approach “M3”
The key here is to identify where your business is under-performing. Data is collected to establish a foundation to frame the right questions & strategies. This intelligence is used to formulate the optimisation strategies.
Tests are designed to challenge the control pages. Constant tracking & performance monitoring are essential. The results are compared to the original strategy to discover what outperformed the control and how it can be further improved.
Documenting the results to enable reference for future tests is key. Leveraging the findings of the tests to other parts of the site spreads out conversion gains. Then repeat the process with new insights!
eConsultancy & RedEye published this framework in their Conversion Rate Optimisation Report.
Measure (seems to be a theme here…)
Taking stock of all the essential elements of the buisness from sources like analytics, surveys & industry research to understand the bearing of the business. This then measures where the business rates in terms of goal achievement & direction.
Once the objectives are understood, gaps are identified by looking at things from the customers view. What are the pain points? You’ll learn more about your customers behaviour and learn how to get them to respond to your call to action.
Based on the gaps, prioritise the tests based on largest impact, cost & time. Then throw on the lab coat, experiment & record the results.
Get those winning changes from your test onto the website. Look to see what works for different audience segments and start the cycle again.
Paddy Morgan from Distilled shared their process.
Discovery – 2 Steps
Step 1 is about gathering the data from both the company, the customer and the website. Touching on Goals of the company, product problems and user experience.
Step 2 looks at the who, what & where of testing. Starting with the big issues first, then fine tuning to the smaller ones.
Experiment Phase – 2 Steps
Step 3 wireframing to ensure the changes are actually testing the hypothesis, are the changes staying on brand and are the changes feasible.
Step 4 implements the design. Tools like Optimizely Unbounce or Visual Website Optimizer are fast ways to create landing pages fast. Always a good idea to compatibility test these sites with all browsers.
Step 5 were you right? Did the test make statistical significance? Make sure you’ve done the test for long enough to get a good amount of data. Find out what worked and what didn’t. Apply it to the site and remember what failed.
That’s the methodology summary finished!
There are a lot of consistent similarities that must be considered essential when considering your own CRO process. You might even call them “pillars“.
Every one gets the data, forms a theory or hypothesis of how to improve, designs & implements the test and records the results. Each time the cycle is completed the next cycle should also be that much more targeted and granular.
It’s a continuous improvement process, which should be making continuous increases in your profit! This is the direction WebSavvy is going, never hesitate to ask us for help. We have analytics training and much more.