We have already mentioned in our earlier posts on conversion optimization process that research and analysis has a great importance within the process. In this post, we will share information about surveys that are important tools in the research stage and what can be done to make effective use of surveys.
First of all, it can be useful to express that it is very easy to determine where the problem is using analysis. However, it is not always easy to find out the issue that causes the problem using analysis. Deeper research can be needed to reveal what is causing the problem. Let’s explain with a short example.
Analytical data can help you discover that users have difficulty with choosing a product. But it may not give clear information on why they cannot choose a product. To discover this, you need to communicate with your users on an individual basis.
In the conversion optimization process, the starting point should be to identify problems with the conversion. When you create a survey, you have to create surveys in parallel with the problem you find out. It will be by a fluke to have an opportunity to optimize your conversions using surveys you will randomly conduct with all users who visit to your website.
Instead, it is critical that you are moving on with an area where you have found an opportunity for conversion optimization. For example, if you have realized that the rate of conversion of users who first visited the computer category within your website is lower than of those who visit the other categories, it would be more appropriate to conduct a survey with the users who have visited the computer category and who have performed certain micro conversions, but have tended to leave without performing macro conversions.
As a result, the first rule for conversion optimization surveys is to find out who your target audience is. Do not start the survey process without knowing the target audience and what you want to learn from this audience.
Now that we have set the target audience, let’s move on to the second stage. How should the survey questions be prepared?
There are two different types of surveys that you can use while creating CRO surveys. The first one is the surveys you want to discover what your problems are, and the second is the statistical surveys you will conduct to discover the accuracy and the validity of the data, or the size of the audience that is concerned with the data.
If you do not have any data related to the reason for conversion problem you have found out (in the example above, it was the low conversion rate of the visitors of the computer category), you have to ask open-ended questions. The number of participants in this survey should be less than in statistical surveys. The responses of the participants should be examined individually and the possible causal data should be found.
But you do not always need it. In some cases, you need to collect data to find out how appropriate the area you are work on statistically. In these tests, you need to reveal how much of your insight reflect the reality for your entire segment by asking users multiple-choice questions, scoring them and ensuring broader participation.
To summarize, surveys are an important tool that we can use to find out the problems of our audience. Before we start using this tool, we must first know why we need a survey. You will not get a result that will allow you to take action from the data you will receive from the general surveys that will be conducted before a problem is detected.
You should go to the stage of preparing the survey questions only after you know who your audience is and why you are conducting the survey. If you do not intend to make comparisons between options and obtain statistical data at this stage, you should ask open-ended questions to users. Instead of asking open-ended questions about future that the user will hypothetically respond to, requesting more realistic data about past experiences and actions they have/have not done recently will provide you with more accurate results.
You can also conduct statistical surveys to test the performance of your findings and the reality factor before making a decision after the research survey. While open-ended surveys require low-participation, you need to reach a broader participation in the tests you want to gain trust statistically.
You can conduct your survey using Hotjar, Typeform, SurveyMonkey, Google Forms, or any tool you can reach. Once you are know your audience and your questions, then it does not make any difference what tool you use.