Data is at the heart of every business. Whenever you make a decision about your customers, you should be basing that decision on the data that you have at hand.
The better armed your company is with data, the better its decision-making. The better the decision-making, the better the outcomes. So what is the best data for your organization to be collecting to enable your organization to become insights driven and help you to make strategic decisions that position your organization ahead of your competitors?
Data is broadly split into two categories: structured and unstructured. Structured data includes close-ended question responses such as age, location and gender and spending habits - the hard numbers stuff.
Traditionally, structured data has been the most common type of data for making business decisions because it is so easy to analyze. But analyzing this structured data alone can put major constraints on an organization’s ability to make strategic decisions and improve their customer’s experience.
When structured data accounts for only 20% of all available data, what is being done with the remaining 80%?
Understanding unstructured data
The remaining 80% of data out there is unstructured. This is where the information is not constrained into clear parameters. Usually taking the form of written information and often referred to as 'open-ended', this data can come from such places as free text responses to customer feedback surveys, social media platforms and customer reviews. Without the limitations associated with structured data, the information contained within unstructured data can vary greatly from one piece to the next.
Due to its lack of constraints, unstructured data is a goldmine for businesses, especially around customer experience. It is a chance for you to immerse yourself in your customer’s thoughts and feelings towards your product or service.
Take the commonly used 'why' question in a CSAT or NPS survey. For a product, this could elicit responses related to the quality of the product, the timeliness of delivery or the user-friendliness of the website. These could be positive, negative, neutral or mixed. But without analyzing the comment and its context, there is no simple way to draw a conclusion on what the customer is trying to express.
Read more: Is tagging / categorization / coding appropriate for analyzing customer feedback?
Up until recently, many organizations' avoided analyzing this unstructured data, as there was no way to efficiently and accurately do so. Organizations that have been analyzing this unstructured customer feedback would do so by manually reading each response and code into an appropriate category – which is a massive undertaking for a large organization with large volumes of survey responses! The other way organization’s have dealt with this unstructured data in the past is to not deal with it all – we know of many organization's who have traditionally ignored the text field responses and focused solely on the structured data.
The power of open-ended customer feedback
Whilst structured questions allow an organization to ask very specific questions, these questions are confined to topics that the organization deems to be important. The problem is that there are other topics that are important to your customers that organizations don’t know about and are not addressing. The power lies in an organization's ability to pull out the important issues captured in the unstructured text field questions in a customer’s own words. When done correctly, this prevents you missing valuable feedback that will assist you to improve your product or service.
It is important to note that structured and unstructured data are not mutually exclusive. In fact, the best outcomes emerge from combining the two. This may be as a means of uncovering the issues that are affecting your highest paying customers by analyzing spend data alongside the unstructured feedback they these customers have provided. It may be linking customer demographic data with their survey responses to drill down on certain customer segments. It is at this junction that you can begin to find actionable insights that will provide direction for your customer experience strategy.
Using open-ended feedback to become insights driven
Given that the benefits are so great, why isn’t every business investing in analyzing their unstructured customer feedback data? Historically, analyzing this data has been difficult, requiring labour-intensive manual coding of text responses. The larger the dataset, the more time and resources required. Additionally, this introduces large variances due to human error and bias: what one analyst considers a key theme of the responses might be totally overlooked by another, leaving the results subjective and often unrepeatable if there is change of staff.
Thankfully, there has been fantastic progress in the computing realm of Natural Language Processing. This enables text analytics software to quickly and repeatedly analyze large chunks of unstructured data, such as that found in NPS, CSAT and customer support data. The growth in this industry has seen it become more affordable than ever before, making it accessible for customer experience and VoC insights teams.
