If your organization is like the majority out there, you will have implemented your NPS surveys and calculated your score and you’re now facing the typical dilemma: what do you do with all those unstructured text-field responses? While it may be tempting to ignore them and focus on the score itself, doing so means you’ll lose valuable information that can help improve your customer experience in an efficient, cost-effective manner. This is how you, or anyone in your team, can find actionable insights in your customer feedback text data.
The core idea of analysing your text data is to understand why your customers are giving you that score. To begin this process of understanding where your business is doing well, and where it’s not doing so well, you need to dive deeper than a score of 0–10. By tracking through the customer responses and looking at what people are talking about, and how they are talking about it, you can begin to identify recurring themes. For instance, you may find that amongst your detractors, the most common phrases are “customer service” and "phone". This will allow you to focus your attention on policies and procedures around phone customer service to improve the experience for your customers, and even convert some of those detractors into promoters.
Your customers are not a homogenous group of people, so knowing all of the different themes people are talking about is just the first step. Once you’ve identified the key themes within your data set, the next step is to segment that data and look at how the differences within your customer group interact with the themes you’ve identified. This will refine the information that you are getting and give stronger insights into what it all means. There are many ways to segment your data and there really is no “best-way”. It is strongly dependent on your business and your customers.
For example, an interesting segmentation is to break apart one-off and repeat customers for an online store. With a much lower cost of returning customers, it can be more efficient to focus on their issues (perhaps your returning customers are frustrated by the checkout not remembering their details). Alternatively, by knowing the concerns of one-off customers (maybe they think your website is difficult to navigate), you may be able to make changes that increases the likelihood of them becoming repeat customers. Another common approach is to segment your customers by age: older customers of a physical shop might be concerned with themes around physical access to the store, while younger customers might be concerned with prices.
Another important note is to account for geographical differences. By segmenting geographically, especially for international businesses, you can identify localised issues and address these within that area, rather than making sweeping business changes. It’s important to note that there’s no limit to the number of segments you can make and that they’re not exclusive. You can combine any number of these to really get inside the minds of your customers.
The last step in finding the actionable insights in your NPS text data is to maintain records of your analysis and review these over time. For a fully implemented NPS program, you should be frequently surveying your customers. As you receive and analyse new data, you can begin to identify and monitor trends over time. Not only will this help you understand the changing focus of your customers, you will also be able to objectively review the effects of any CX projects you have undertaken.
After identifying a segment and a theme and implementing a change in your business to fix the issues you see arising, you need to circle back to your NPS data and confirm that the change has had a concrete impact on your customers. If you make changes to your website to make items easier to find, then a successful intervention will mean that fewer customers talk about this issue in surveys done after the change. If you make a change, but that change isn’t represented in your surveys, then that means the core problem has not been addressed, and the experience of your customers is still falling short of where it could be.
How much time do you spend manually analysing your text data?
Each of these steps can be time-consuming and labour-intensive. It can take a data analyst days, weeks or in some cases months to go through your responses manually and identify common themes. To then segment this, they’ll need to start all over again.
The good news for Data Analysts and Voice of Customer Managers is that collecting and analysing NPS is easy when you have the right tools.
Natural language processing applications can do this in minutes (if not seconds!), allowing you to chop and change the data as much as you like to get a clearer picture of your customers and to quickly identify actionable insights.
So, now is the time to take that mound of survey responses and turn it into a valuable commodity for you organization’s growth!