The secret to uncovering customer insights with AI

Explore

The secret to uncovering customer insights with AI
The secret to uncovering customer insights with AI

An inside look at how Dstillery uses AI to uncover impactful insights from customer interviews

An inside look at how Dstillery uses AI to uncover impactful insights from customer interviews

Taejin In

As the SVP of Product at Dstillery, I’ve managed and worked alongside dozens of Product Managers throughout my product career. When I encounter successful PMs, I’ve always noticed one key characteristic — the best PMs are masterful at talking to customers and extracting meaningful insights. These insights are then used to shape everything from small features all the way to the company’s vision for the future.

With how important this skill is within a PM’s arsenal of tools, I’ve been surprised with the lack of discipline when it comes to managing customer interviews and extracting insights. From poorly thought out questions that leads to biased responses, to important feedback getting lost in the shuffle, it is absolutely critical that PMs approach the customer interview process with as much rigor as they do with the product development process.

Have a centralized source of truth.

Nothing is more frustrating than remembering that a customer said something that validates or refutes a product decision, but not being able to find the interview or the direct quote. From in-market conversations from sales calls, to interviews driven by Product Marketing in preparation for a product launch, to interviews with customers to start defining requirements, there is no shortage of places where customer feedback is captured. By having one place where all these engagements are centralized, you’ll know exactly where you need to go to find that piece of evidence to inform your product development efforts.

A good question is half the battle.

You know when you’ve asked the right question. You’ll hear things like “that’s a great question”, “that was very well put”, “I haven’t thought about it like that”. Asking the right questions is how you will discover the real gold — responses that refute or validate major product decisions; responses that make you think about the problem in whole new ways. The great thing is, these situations can be intentionally designed by spending a lot of time making sure you get the questions just right. Asking questions that are biased, leading, or non-actionable, is a waste of everybody’s time.

Be intentional about participants.

Dstillery’s primary customers are media agencies and there are big differences between ‘hands-on-keyboard’ operational people, and people who manage a team of these people. People may be in similar teams, but can be motivated and incentivized by completely different things. When crafting and defining interview questions, it’s critical to understand things like: are you talking to the person who would be using your product, or the person who will be approving the use of your product? Has the person engaged with your product in the past? Has the person spoken about this type of product publicly in articles or social media posts? The more you understand about the participant before the interview, the more insightful the conversation will be.

The Solution: Dstillery’s Customer Interview doc.

It’s easy to send a document of best practices to PMs on your team, but how can we increase the chance of success with every PM and every customer interview opportunity? That’s where this Coda doc comes in. We built four sections to specifically address successfully executing customer interviews and extracting insights.

Interview Guide

Creating a good question is half the battle. First, I suggest creating a list of ‘things you want to learn’ in the top of the document as your team brainstorms questions. This will ensure that the questions circulate around the ultimate goal you are trying to reach.

AI use case: When you type in the ‘things you want to learn’, Coda AI automatically populates 3 questions that could help you answer the thing you were hoping to learn.

As you and your team start adding new interview questions, people can simply +1 any questions in the ‘Thumbs up’ column to show their support for questions they want included in the interview. Also, when crafting questions, it’s often hard to see if there are any hidden biases within the question. The ‘potential bias’ column allows others to indicate that the question might be potentially biased and should be reworded or reconsidered.

⁠

Once the set number of questions are identified (this can vary based on whether it is a 30 or 60 minute interview), you can finalize the list and move on to the next step.

Participants

Having a single place to align on the participants of the interview is important. Besides understanding roles and company, have a place where people can share notes about the potential interview candidate. Notes can include things like:

Have we had engagements with this person in the past?

Are there any articles, publications, social posts, this person has posted?

Has this person’s team purchased our products in the past?

Beyond using this page as a place to jot down information about the person you are interviewing, this page can also be used as a centralized place for scheduling information.

Interview notes

As you run the interviews, it’s important to have one place where all the notes on a participant’s responses are captured.

Side note: Remember to listen intently as customers respond to your questions. Active listening will lead to you asking the right follow-up questions, and responses to these follow-ups often lead to really insightful information.

⁠

With dozens of interviews multiplied by dozens of questions, it’s easy to get disorganized and miss important information. It is super critical to stay organized in this step.

Insights

After the interviews, the hard work is done and now comes the fun part. Go through your notes and extract common themes and insights that can inform your product development efforts. While this part can be very time consuming, using Coda AI, you can instantly see what a LLM (large language model) uncovers from the responses.

AI use case: All your customer responses from the interview questions will be automatically summarized in the ‘Coda AI Insights’ column.

⁠

Not only does this tool save your team significant time, I’ve often seen the AI-based results sparking new ideas that don’t come up during team conversations.

Ready to get started?

Press the button below, copy this template, and get ready to completely revamp the way you run customer interviews and extract impactful insights!

⁠

Copy template

⁠

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.

The secret to uncovering customer insights with AIThe secret to uncovering customer insights with AI