Turning survey data into insights

Turning survey data into insights

What’s the easiest way to collect customer feedback at scale? The humble survey.

Surveys are incredibly simple to create and send, whether you have ten customers or ten million. Survey tools are everywhere, and it seems like every month they get even easier to use and deploy.

Everybody knows that businesses should collect customer feedback, so this is all good news.

But what do you do with all that survey data? How do you use it? How should you think about the different types of questions on your survey and the different types of data they give you?

There are plenty of ways to answer these questions, but here’s a simple way to think about it:

Surveys give you two types of data, score data and amplification data.

Survey score data

Survey score data is data that’s essentially a key performance indicator (KPI). It’s quantifiable data and usually gives you an overarching number to focus on. Common examples of score data include Net Promoter Score (NPS), customer satisfaction (CSAT), and Customer Effort Score (CES).

In short, score data is the type of data that normally shows up on your customer insights dashboards. It’s high-level, top-line data that gives you info on how your team and company are doing at a quick glance.

Survey score date is particularly useful because it’s easily digestible. You don’t need to be a CX professional or an analytics wizard to understand that if NPS was 70 last quarter and it’s 20 this quarter, something’s probably gone very wrong. Sharp decreases in score data typically mean some kind of intervention is needed—and the sooner, the better.

Another valuable aspect of score data is that it can be helpful in setting North Star goals.

North Star goals are long-term, aspirational goals that are meant to set the direction and trajectory for your company (thus the name). If your leadership team wants to prioritize becoming more customer-centric over the next few years, it’s relatively easy to set a North Star goal that you’re going to increase NPS by 10 points each year.

Achieving big goals is easier said than done, but score data can play an important part in functioning as a North Star goal for your organization.

Flowchart showing the relationship of survey data, score data, amplification data, attribute data, and verbatims.

Survey amplification data

The second type of data your surveys give you is amplification data.

Think about a guitar amp. If you’re not musically inclined (guilty!), an amp works by picking up the signal from your guitar, feeding it through a circuit to “shape” the sound, then amplifying it and pushing it out through your speakers. The amp takes the relatively weak signal from the guitar and makes it louder and fuller so that your audience can enjoy it.

Amplification data works the same way. Certain questions on your survey will provide information that “amplifies” your score data. It paints a fuller picture of what your customers think and feel, so you can enjoy it—or at least understand it—more fully.

In other words, amplification data helps you understand why your score data is what it is and how to improve it moving forward.

There are two key types of amplification data: attribute data and verbatims.

Attribute data in customer surveys

Attribute data is usually quantitative, similar to score data. The difference between the two is that attribute data normally falls underneath your top-line score data.

For example, say CSAT is the top-level KPI in your survey data. You ask customers “How would you rate your overall satisfaction with your customer support experience?” They respond on a scale of 1-7, and you use that info to calculate your CSAT score.

After customers answer that first question, you might ask them additional questions about specific aspects of their experience, such as:

  • How would you rate the responsiveness of our support team?

  • How many times did you need to contact us to get your issue resolved?

  • How easy was it to get the help you needed?

As customers answer these questions, it paints a fuller picture of their experience. Your overall CSAT score might be 85%, but it’s these follow up questions and the resulting attribute data that helps you understand why customers are so satisfied with your support experience.

It’s important to note that any of these attribute questions could become your top-level KPI. You aren’t limited to NPS, CSAT, CES, or any other trending acronym. If you discover that the single most important thing for your customers is a speedy response, you could make the first question above (How would you rate the responsiveness of our support team?) your top-level question and the source of your score data.

While score data is often valuable for keeping your whole organization informed and aligned, attribute data is usually most helpful for your CX team. Because it’s more specific and breaks down each customer’s experience into more granular categories, attribute data is often useful for agent- or team-specific coaching and improvement efforts. As such, your support team leads and managers should probably keep a close eye on this data.

Attribute data helps you understand what went right and what went wrong during a specific interaction.

There’s one other big implication of attribute data: improvements to each individual attribute’s area should lead to an overall improvement in your top-level score data.

The easiest way to drive positive improvements is to categorize your responses and look for trends across your different attribute questions.

For example, say most of your NPS detractors are saying it takes too long to get a response. If these are your most negative overall responses and the majority of them also seem to agree that it’s taking too long, then it should seem pretty obvious: solving for this complaint ought to lead to a decrease in detractors and an overall improvement in your NPS.

Analyzing attribute data in this way gives you actionable insights that you can use to make smarter business decisions.

Verbatims in survey data

Verbatims are the second type of survey amplification data. Your verbatims are your open-ended responses from customers. Structured questions (like multiple choice) give you nice clean data; verbatims let you hear from your customers in their own words.

Structured questions (like multiple choice) give you nice clean data; verbatims let you hear from your customers in their own words.

Verbatim data typically comes from your free text survey questions, such as “What else could we do to improve your experience?” or “Why did you choose that response?”

If you’re an insights or support manager you probably already know that analyzing verbatims can be a painful and slow process. That doesn’t have to be the case though. When you’ve got a bunch of customer verbatims, you have two possible approaches to analysis: manual analysis or using a feedback analytics tool.

Manual analysis of open-ended survey data

Manual analysis of open-ended responses is hard because they aren’t binary and aren’t directly quantitative. They’re all over the map (just like your customers!).

There’s no getting around the fact that manual analysis involves a ton of time. When you’re having a human manually read and categorize hundreds or thousands of data points, it’s going to take a while.

Here’s the best way to do it:

  1. Categorize responses by theme and sentiment from the customer’s perspective. You might include a few different categories, such as the product and the aspect of the experience the feedback pertains to. Categorizing the sentiment at the same time is also helpful, because if 100 customers mention your support’s responsiveness, you’ll want to understand at a glance what percentage of those comments were positive, neutral, or negative.

  2. Count up your responses by category (and sentiment, if applicable). When you’re sharing reports or updates with people from across your organization, including direct quotes from customers can be powerful. But you’ll also want to try and quantify your customer’s verbatims, because it’s a huge help in prioritizing future decisions.

    One way quantifying verbatims is useful is that it helps you avoid “squeaky wheel syndrome,” where you’re making decisions based upon the loudest or most negative customer comments.

    For example, say you have three furious customers complaining angrily about your tool’s reporting features and three hundred complaining less angrily about your clunky UI. It may be tempting to fix the reporting features because those customers are so outraged—maybe they’re even contacting your CEO directly—but the survey data clearly shows that focusing on UI improvements will have a bigger impact.

Make sure you’re approaching this categorization process from the customer’s viewpoint, not your own. It’s tempting to categorize feedback based on your knowledge of your internal processes, existing priorities, or biases. While everyone feels this temptation, giving in to it can inaccurately color your understanding of your customers’ needs and priorities.

Using a feedback analytics tool for open-ended survey data

The second—and far easier way—to analyze open-ended survey data is to use a feedback analytics tool like Kapiche. Feedback analysis tools aren’t limited to open-ended data, but this is one area where they can really shine.

Watch this demo video to learn how Kapiche analyzes feedback data.

Feedback analytics tools are able to process unstructured text data far faster than any human—in minutes, instead of hours, weeks, or months. These tools are unbiased, are better at identifying themes, and can help you identify emerging trends in your customer survey data.

Getting started with a feedback analysis tool is often simple and easy:

  1. Pick your tool

  2. Connect your data sources

  3. Get to work

When your tool takes care of the time-consuming analysis, it frees you up to focus on understanding the insights in your survey data so that you can give deeper insights to your organization and better serve your customers.

Survey data alone doesn’t get you anywhere

Many businesses get in the habit of running regular customer surveys and yet get very little value from them. That’s infuriating for customers—because they feel ignored—and demoralizing for your customer experience and insights teams.

At Kapiche, we’ve worked with customers who have hundreds of thousands of survey data points sitting in a spreadsheet, just waiting to be processed. Due to the massive time commitment such a project would require, no one ever did anything with them. Those survey responses never led to any impactful changes for those businesses.

There’s no point in running customer surveys if you can’t derive any actionable insights from the results. If you’re struggling to turn your survey data into valuable insights, you can transform your customer experience by using Kapiche.

Check out a demo today and convert your survey data into instant insights.

Share to: