Did your latest A/B test or website redesign not have the result you envisaged? Were you sure that you knew what was wrong with the page, but can’t understand why the change hasn’t positively impacted the performance metrics? You’re not alone. When it comes to deciding where to focus your efforts, many businesses are facing the same challenges and approaching website optimisation in an unstructured and haphazard manner. The following are some of the most common ways we see businesses making important decisions;
Table of contents:
- 1 Designing by committee
- 2 The hippo is making the decisions
- 3 Copying a competitor
- 4 Why understanding customer needs will improve your conversion rate
- 5 What is a hypothesis?
- 6 How to generate the best ideas to test?
Designing by committee
Your designer is trying to appease a group of stakeholders representing their own agenda. There is no common vision for the design and it has most likely not been created with the end customer in mind.
The hippo is making the decisions
A common problem is when the hippo (Highest Paid Person’s Opinion) starts making uninformed decisions based on ‘instinct’ and in line with a strategy they implemented; this is basically just their opinion.
Copying a competitor
You’ve seen a new widget on a competitor site, so you presume it’s a guaranteed win. You decide to start building an even bigger and better version.
Why understanding customer needs will improve your conversion rate
All of the above approaches are missing one important influence – the customer. You are not your customer and neither is your competitor. Not to mention the “internal blinkers effect” where you use the site so much that you are almost on autopilot. Meaning you can’t see the obstacles that are acting as barriers to your customers as you’ve become immune to them. You also can’t guess what customers might like to see on a page, their considerations during a purchasing decision, or their motivations to buy.
Like many businesses, you probably don’t have unlimited traffic or an abundance of design and development resources, so you have to make every test count. Data-driven decision-making with customers at the heart of your strategy will lead to informed ideas providing you with a higher % of your tests and improving conversion rates.
Let’s start at the beginning; every test needs a hypothesis. Hypothesis development is one of the most important stages of a testing cycle. Without a good hypothesis, the test is already hindered. A hypothesis should be written as the very first step before any wireframes, designs, or solution decisions have been made.
What is a hypothesis?
By changing X to Y, it will cause Z (effect)
As an example, changing the registration form from three pages to one will increase the number of sign-ups. When it comes to the result of the test, the hypothesis is either proved or disproved; there is no middle ground.
A hypothesis is usually accompanied by a rationale. The rationale helps to support the hypothesis by including more information as to where the hypothesis has come from, what the problem is, the impact it’s having on the customer, and how the change will address this.
How to generate the best ideas to test?
We’ve covered how to write the hypothesis, but where do the ideas come from to feed into the hypotheses? Here are a few options that we tend to use:
Ask your customers
- Running remote or moderated user research sessions is a guaranteed way to generate a list of hypotheses and find out what is important to your customers.
- Top tip: Set up a mix of open and focused scenarios, so that you can gain an understanding of their thought process and what they would naturally do, but also uncover obstacles on key journeys.
- Onsite survey tools such as 4Q or Qualaroo are quick and easy to use which enables you to start understanding customer intent whilst they are on your site.
Review your data
- What is the data telling you about visitor behaviour? Look at your funnel and assess where visitors are dropping out. Is there an obvious problem area where you should start?
- Where possible, validate the qualitative research insights with quantitative data to further strengthen the hypothesis. This will help to ensure you concentrate your tests on the critical problem areas.
Talk to your customer services team
- Your customer service team has a first-hand relationship with customers. They talk to them daily and will most likely have a good understanding of common customer frustrations. They are a source of customer knowledge; use them.
This method of hypothesis development ensures there is valid insight feeding into the hypothesis, thus the test will both, a) target a real customer issue, and b) also gain further customer insight through testing a variation that has been designed to change behaviour.
- Every test should have a hypothesis
- The hypothesis should be formed and supported by customer insight and data
- Optimise for your customer, not yourself
- Click here to learn more about Conversion Rate Optimisation