Interpreting Data and Understanding your Findings

While it may sound obvious, interpreting data correctly and effectively can be quite challenging, and you may unintentionally misinterpret your findings. Conducting your own customer research can result in bias, both in survey design and data interpretation. Ultimately, you want to see positive results - but this can lead to unintentionally leading questions, or misunderstanding of participants' responses. To overcome common challenges, this article offers some tips and tricks to help you construct effective research surveys, to obtain accurate, useful data you can use to improve your brand and services.

Interpreting data accurately and effectively

There are several things you can do to try to ensure that your data is being collected and interpreted accurately and effectively, to provide you with the best opportunity to use your findings to grow and improve.

1. Allow negativity

It can be all too easy to find yourself interpreting data optimistically. Focusing on positive comments, dismissing negative scores as anomalies; while you may not do it intentionally, wanting the best for your business can bias your interpretations. Remember that negative comments are the best way to learn and improve, as they can help you identify the areas that need the most work. Look for correlations in negative scores and comments, and you may just discover the core of the problem.

2. Conduct multiple types of surveys

You might also find it beneficial to examine multiple facets of the customer experience. A customer satisfaction survey can be very useful to know if your customers are happy with the services you have provided, but undertaking further research (such as a customer experience or customer perception survey) might help you to gain a more complete image of your customers' needs, as well as your brand image and position in the marketplace.

3. Conduct regular, repeated research

It is vital that the customer research you undertake is not analysed and then promptly forgotten. Repeating research regularly allows you to compare data over time to identify trends, and helps you to more quickly spot areas for improvement. Regular research can also improve the accuracy of your findings, by reaffirming them at a later date. We've written an article on how often you should conduct customer surveys if you're not sure.

4. Think of your data as interlinked

Each question has more meaning when you place it in context with the other questions in your survey. Perhaps your customers love how easy to navigate your website is, but they feel that it could use some more variety in content. As a result, they may rate your website quite poorly overall, but looking at the rest of the data informs you that the only area that needs work is the type of content available. Taking questions in isolation of the rest of the survey can impact how effectively you are able to interpret the reasoning and meanings behind each response.

Why is context important?

One of the biggest mistakes we often see is when research findings are taken out of context. While statistics drawn from your survey can make for very appealing headlines and infographics, they were never intended to be isolated from the rest of the findings from your research. A percentage or figure can only tell you so much if it is not supported with further information, especially when it comes to people's views and opinions. An '8' on a scale of 1-10 can mean very different things to different people, so ratings should always be viewed in context; while they can be useful to quantify something that is otherwise unquantifiable (such as an opinion), and to provide a general idea of where people's perceptions lie, they cannot give you the whole truth. Furthermore, the context of the research itself is also crucial. An average rating of 8 out of 10 may sound great, but if this data was gathered from a sample of 12 participants, it may not be representative of your customer base as a whole.

Wherever you intend to utilise your findings, make sure you back them up with details about your survey (population type, sample size, etc.), or include supporting comments - a figure without any context can be misleading.

Using Mixed Methods to support your findings

One of the most effective ways you can make sure you understand your data is to back up your quantitative questions with open-ended, long-answer questions. Gathering a Net Promoter Score can help to quantify your customers' overall perception of, and loyalty to, your brand, but what then? Knowing your NPS is -36 is useful as it informs you that your customers are less likely to recommend your brand, but it does not tell you why. Following up your NPS question with an open 'comments' space can allow your customers to provide a reason for their answer, and help you to understand how you can improve. This logic can, and should, be applied to most of your quantitative research questions. For more information on using NPS appropriately, take a look at our 'Net Promoter Score vs Customer Satisfaction' article.

On the other hand, make sure you offer plenty of response options for quantitative questions. If you would like to hear how your customers would rate your customer service team on a scale of 1-5, the best way to ask would be to provide multiple sliders with different criteria. For example, perhaps you ask participants to rate your customer service team on friendliness, professionalism, and ability to provide the help they need. This can help you to identify exactly what areas need work; maybe your customer service team are kind and friendly, but a little too much so, and they lack professionalism. Or they may present themselves very professionally, but struggle to provide the support your customers are looking for.

Do not focus in on one type of question. Not only do mixed methods help to support your findings, but they also break up the monotony of the survey, reducing the risk of survey fatigue which could cause respondents to give up midway to completion, or to rush their answers. A column of 20 Likert scale questions could be very off-putting to participants, but if these 20 scales are broken up by open-ended questions and checklists, participants may stay interested for longer - as well as providing more useful data for you to analyse.

What can I do to make sure I am interpreting data correctly?

From the first stage of survey planning, right through to the analysis of your findings, you can take precautions to ensure that your data is useful and accurate.

To begin with, do not rush your survey design. Make sure that you are asking the appropriate questions and offering the right response options. Use different types of questions (both quantitative and qualitative) and ensure that every question has the potential for negative, and positive, feedback. Using a range of questions allows you to understand the reasons and motivations behind your customers' responses, as well as allowing you to compare related responses across multiple questions.

When you are interpreting data, acknowledge that you may receive some criticism or negative feedback. Accept that no business is perfect, and know that you may receive complaints or requests for change. Be prepared to take this feedback constructively, using it to improve your services and increase customer satisfaction and loyalty.

You can ensure your data is interpreted accurately, through an objective, unbiased lens, by using an external market research organisation. At Mackman Research, we offer a range of research services and will provide you with professional, accurate, and useful findings, to help your business achieve its full potential. Contact us to find out what we can do for you.

By Jess Crago