4 ways product managers can use qualitative data in customer feedback


Why customer interviews and focus groups aren’t enough


1. Use the volumes of customer feedback data that already exist


2. Get insights from a variety of unprompted feedback


3. Use tags and sentiment to automate text analysis


4. Put existing qualitative data to use in a data-driven way


Take a data-driven approach to analyzing qualitative feedback

Try us for free

As a product manager, you aim to be data driven in your product decisions and you want customer insights to inform your strategy. When converted to data, it turns out qualitative customer feedback can help you excel at your job—by providing valuable insight for product discovery, product strategy, and launch delivery.


Why customer interviews and focus groups aren’t enough

You make multiple product strategy decisions a day and can’t always ask customers what they think or how well a new feature solves—or doesn’t solve—their problem. You’re afraid to annoy your customers with too many questions; even your “friendlies” have stopped responding to your Slack messages. 

Focus groups are an additional resource for exploring new ideas but it’s often hard for people to guess how they might use a product in theory.

Without the big picture about what frustrates, delights, confuses, annoys, or excites actual customers of your product, you can’t prioritize or iterate with great confidence.

If you could just tap into customer insights without having to ask your best customers, yet again, for feedback.

Here are four ways to look beyond focus groups and customer interviews to round out your user research.


1. Use the volumes of customer feedback data that already exist

Though focus groups and one-on-one interviews are the gold standard of user research, they take time and significant resources. Conducting focus group research in particular can also be prohibitively expensive. 

What other sources of customer insight can you turn to?

Helpdesk tickets, online reviews, social media interactions, and free text feedback found in net promoter (NPS) and customer satisfaction surveys (CSAT) add up to substantial first party data from your customers about their user experience. Most companies already get this data daily from existing channels. There’s no need for an extensive set up or additional expense.

Insights from these datasets are complementary to traditional user research. They can highlight new issues or ideas you hadn’t thought to cover in your focus groups or customer interviews. 

Best of all, the data is always fresh because it’s collected in real time and continuously. 


2. Get insights from a variety of unprompted feedback

Helpdesk tickets, online reviews, and social media interactions are great sources of real feedback because you’re not leading the conversation with questions about a particular topic. Instead, the customer is taking the lead. Direct feedback from users who have already interacted with your product can provide more useful signals than users estimating how they may feel about your product.

NPS surveys and CSAT surveys also give customers an opportunity to share what’s on their minds without restricting feedback to specific topics.

What about the highly vocal users who have a lot of (often colorful) opinions? Do they skew the results and overrepresent a particular point of view? They can, if considered in isolation. Bringing feedback together from multiple sources helps to mitigate the “squeaky wheel” effect. 

To stay focused on what truly matters, segment your analysis by customer type across your data sources if possible; you’ll be able to spot the issues that have the most business impact.


3. Use tags and sentiment to automate text analysis

You have gained access to all these different sources of text-based feedback. What then? Instead of reading through each one individually, use text analysis in an automated way.

Tagging and sorting free-form text from customers helps you quantify your qualitative data: by sentiment, emotion, benefit, product type, feature request, user flows, and other categories. You can use keywords or themes to categorize. Lots of customer support tools offer automatic keyword tagging. You can even do this at scale automatically across your various customer feedback sources.

Use the volume of topics mentioned to assess impact. For instance, 2,537 mentions of checkout probably warrant more urgent attention than, say, 29 mentions of background image. Particularly if sentiment analysis tells you the majority of the checkout mentions are categorized as negative. It could indicate users are having trouble completing a purchase at checkout.

The source of feedback also helps you understand context, urgency, and level of effort needed to address it. Urgent issues that appear in helpdesk tickets may require more immediate but short-term attention than new feature ideas that came in through a survey. But it doesn’t mean new feature ideas aren’t important. Knowing the volume of each topic in the customer feedback helps you identify what needs priority; you can then plan your time accordingly for strategic work that increases adoption, deepens user engagement, or expands reach into new markets.


4. Put existing qualitative data to use in a data-driven way

Once you’ve embraced multiple sources of feedback for product strategy, how do you put it all to use? 

  • Compile it. There are lots of qualitative data sources that need to be brought together. Figure out the workflows that will automate this for you. Aggregate data from different sources by connecting apps. Tools like Zapier can help simplify this.
  • Normalize it. You can easily get bogged down quantifying and analyzing each dataset in silos. Normalize the qualitative data across a few key dimensions—for instance across topic mentions, sentiment analysis, or emotion—to make your efforts as data-driven as possible.
  • Count and track it over time. Every product manager loves to spot trends. Once you’ve decided what dimensions to use for normalizing free-form text, it’s easier to choose how to count data. By 30-day, weekly, daily, or hourly, depending on what makes sense for your business. Choose a few metrics that will get you the greatest amount of insights instead of too many metrics that could overwhelm your efforts.
  • Automate it. You don’t have to saddle an analyst or associate product manager with all this work. Automated workflows are increasingly the norm. Beyond automating the aggregation of qualitative data, you can also automate the analysis of it. (Automated text analysis is exactly what Viable was designed to do.)

    The more steps you can automate between the moment feedback comes in and the insights you get out of it, the more time you have to focus on building better products.


Take a data-driven approach to analyzing qualitative feedback

Viable's advanced language AI automates tagging and sorting of open ended responses from your user research surveys and other data sources. If it saved you at least few hours a week, why not check it out?

Try us out for 30 days, no obligation. Analyze up to 50,000 pieces of data (or 30 days worth). Connecting data sources takes just a few minutes. Then you can start asking questions and getting answers in seconds. What do you have to lose?

Try us for free

Viable Team

Viable Team

Staff

Last Updated: 03/01/21

Related walkthroughs

We raised funding to automate qualitative analysis!

SAN FRANCISCO, CALIFORNIA (September 16, 2021) -- Viable, an analytics startup using artificial intelligence to automate text analysis of customer feedback, announced it has raised $3.9 million in pre-seed and seed funding to bring to market its next generation qualitative analysis AI tool. Craft …

Viable Team

Viable Team

Staff

Churn analysis: a practical guide

A leaky bucket in business lingo means you’re losing customers. Every company has some customer churn. A significant amount of churn can be a sign of problems with the product or the customer experience—or it could be due to changing market trends. By conducting customer churn analysis regularly, …

Viable Team

Viable Team

Staff

Why we built a GPT-3 app

GPT-3 is an advanced AI system that produces natural language text by predicting what comes next in a text sequence. It’s one of the largest neural networks available, with 175 billion parameters. GPT-3 was trained with large amounts of information from the internet.  Thanks to all that …

Viable Team

Viable Team

Staff