Today, we will take you through the three steps required to evaluate your open texts survey responses with the help of Caplena, the market-leading text analysis tool!
“What Features/Practices Annoy You The Most About Apps/Websites?”
We asked 800 contestants this question to demonstrate how Caplena can analyze survey responses. In doing so, you will learn how to employ this approach to analyze your own survey data! Are you ready? Let’s go!
The basic principle when having to evaluate free-text (e.g. survey responses) is called Coding: The process of assigning one or more tags or topics to every text element.
When you have assigned codes to a majority or even all of your text elements, you can create statistics, charts, and dashboards, describing your results:
Did you know the top three things people find annoying about websites are ads, pop-ups, and lagginess? You can check out the interactive dashboard we made for this blog post here.
Of course, these steps can be done manually in Excel or by using some other coding software. It even should be done there if you have, say, only 50 text elements. However, this is totally impractical with 200 – and borderline impossible with over 5,000 responses.
Caplena will eliminate much of the repetitive work in analyzing open-ends. Spend more time on the task which requires your insights instead: Making sense of the results.
The first step is obviously getting your hands on the data. There are several ways to accomplish this: Ask your clients for feedback, send out an NPS survey or scrape the web for social media data posts or product reviews.
We support various file types, such as Excel, CSV, or SPSS. You also have the choice to directly input your data in the form of numbers, dates, booleans, strings, NPS and CSAT scores, etc. Using our latest integrations feature, you can also simply copy and paste links of survey-type websites, and download the respective responses like that. See more about our integration features here.
After completing this step, a codebook is generated, i.e. a list of codes (tags) that should be attached to the texts. Keywords, as other tools deliver them, might be Service or Price.
You can also upload a codebook you may have from similar or identical historical surveys so that you don’t need to start from scratch building a new one. We also offer a wide range of survey codebook templates that you can select and then adjust to your specific needs.
The above is an example of a codebook and its categories. A general category, such as Customer Service, usually has codes within it, like Good Support, Bad Support, and so forth.
The survey we conducted had a number of responses from which Caplena identified the relevant codes to connect to different parts of the response. This specific screenshot was made in the “Fine-tuning” view, which is completely optional. Nevertheless, we recommend looking through a few responses to see if the AI added the right codes, and if necessary, deleting and adding the correct codes.
The process on Caplena is a prime example of augmented intelligence: Our algorithm detects topics, but your industry knowledge and abstraction capabilities are required to improve organization & clustering.
We can now begin to make sense of the survey responses after putting the unstructured text into structured categories. Caplena provides a visualization module, in which you can easily create charts and dashboards, which you can either download or share with your team/clients directly through a shareable link.
It is also possible to download all the raw results – the charts themselves can also be exported for use in their visualization toolkit.
Check out this dashboard we made.
It took us less than 15 minutes to create!
The mechanics of creating bar, pie, line, treemap, or graph charts on Caplena are very simple, but making sense of your data is not. Getting to the essence of the data and telling a compelling story with a few words and a few charts is the most critical aspect of any analysis. This is also where you should invest most of your energy.
Within under half an hour, we analyzed our survey responses and have the data ready for presentation on Caplena’s interactive dashboard. So what have we learned? It is all about taking advantage of today’s technologies in the most efficient manner and spending your time on the things that machine learning cannot automate – such as making sense of the results.
How do you decide on what kind of survey to use? We’re eliminating some of the guesswork for you by giving you foundational advice to creating an effective customer survey, and a good old-fashioned comparison between open-ended vs. multiple-choice.
Customer feedback is about more than knowing whether your customers give your product three stars or five stars (although that is a valuable insight). It’s also about gathering the information that can help you make critical decisions and improvements….