How to use Generative AI to build better products with customer feedback

This is a recap from our conversation with Shyvee Shi on the topic of how to use Generative AI to build better products with customer feedback.

Shyvee: What initial spark or problem led you to start Enterpret, and how did your background prepare you for this venture?

Varun: The idea for Enterpret was inspired by my experience building out the customer success team at Amplitude, where I led hundreds of deployments of product analytics for small consumer apps to large Fortune 500 companies.

I witnessed the evolution of product analytics as a category and, more importantly, observed a critical gap in understanding the 'why' behind product analytics. Teams can track retention and their conversion funnel metrics to the decimal, but if you ask the simple question, “What are the top three customer pain points?” Crickets. Regardless of size or industry, most teams lack a clear understanding of customer pain points or actionable insights.

Enterpret began to take shape in 2020 after I had worked on NLP projects at Scale.ai. My brother Arnav and I started exchanging ideas about how to solve the gap in product analytics using the rapid advancements in AI. Arnav’s background is in computational linguistics and NLP research, and he led the Engineering team at Uber, so he was very much at the forefront of research and development in the space. Together, we’ve embarked on this venture to decode customer feedback into actionable insights for product development using AI.

Shyvee: Can you share some real-life case studies from companies that have used Enterpret to empower their product organizations?

Varun: We’re lucky to work with some of the most customer-centric companies in the world, like Canva, Notion, Loom, and Apollo.io. Here are a couple of ways we’re helping these teams:

Notion Improves Product Load Times: Notion uses Enterpret to run its Voice of Customer initiatives to help prioritize complaints. Enterpret helped to identify the top issue around slow product load times. After they shipped these enhancements, Enterpret helped verify a 50% improvement in load time and a 65% reduction in related complaints, illustrating the impact of combining quantitative and qualitative feedback for product optimization. These improvements enhanced user satisfaction and validated the effectiveness of the changes through direct customer feedback.

Apollo.io connects feedback revenue impact to drive their products forward Apollo.io uses Enterpret to consolidate all feedback and customer interactions to align the organization around the Voice of Customer. These VoC insights help Apollo.io to identify the Pareto, which is 20% of the issues that cause 80% of the damage or 80% of the inbound contacts. It has empowered the team to leverage their engineering resources to ship valuable features while minimizing bug fixes and" keep the lights on" work. The team has seen a 40% reduction in support issues around themes they’ve chosen to address.

Shyvee: Can you share your biggest challenges while building Enterpret and how you overcame them?

Varun: The biggest challenge was altering the perception of customer feedback from a procedural task to a critical, trackable dataset. We tackled this by demonstrating the tangible benefits of actionable insights derived from feedback analysis, emphasizing the importance of integrating customer perspectives into the analytics framework.

Another challenge was ensuring the adaptability of our AI models to the fast-evolving nature of customer feedback and product development. Continuous innovation and a flexible approach to model training are helping us stay ahead.

Shyvee: What are some unique considerations working with generative AI?

Varun: Working with generative AI requires meticulous attention to the format and quality of customer interaction data. We've developed processes to clean and standardize data, removing irrelevant patterns and ensuring consistency. This foundation allows us to leverage AI to generate granular and actionable insights. Additionally, we focus on natural language interactions, enabling users to query our database effortlessly, thus making feedback analysis an integral part of the product development cycle.

Shyvee: How does Enterpret position itself to stand out? What are some “moats” Enterpret is building for long-term sustainability?

Varun: Enterpret distinguishes itself by providing granular, actionable AI models tailored to each customer's unique needs. Every Enterpret customer has a custom taxonomy built to reflect their customers, product, and business. Our unique AI models can adapt to each customer's specific feedback landscape.

Unlike generic solutions, our "custom-fit tailored suit" approach ensures that insights are highly relevant and actionable, making us a preferred choice for product and customer experience organizations seeking to base decisions on deep, actionable customer insights.

We are also deeply committed to making the feedback analysis process seamless. By continuously investing in AI and ensuring our models remain at the cutting edge, we create a competitive advantage that is difficult for new entrants to replicate quickly.

Most importantly, when it comes to learning and acting on customer feedback, it boils down to, "What is the revenue impact of improving this thing, and is this the highest leverage investment we can make?” By keeping a live mapping of all of your customer feedback across all sources to the system of record for user/account/revenue data, we enable product teams to answer questions that matter. Enterpret is connecting feedback to revenue impact to inform critical decision-making. See it in action.

Shyvee: How do you envision the future of product development changing in the next 3 years, and what role do you hope to see Enterpret play?

Varun: As digital transformation accelerates with the emergence of generative AI, I see product development becoming increasingly customer-centric. Feedback analysis will continue to play a more pivotal role in shaping products.

Enterpret aims to be at the forefront, enabling companies to understand and act on customer feedback at scale, thus ensuring products meet and exceed customer expectations.

Shyvee: As you reflect on the journey so far what has been the most pivotal or significant learnings during your journey building Enterpret?

Varun: I have always believed in building a team of people united in their shared conviction that a problem is worth solving.

The second is to stay humble. I look for people with a humble mindset and a strong desire to learn and improve.

Reflecting on the journey, these two principles of great people and humility have guided Enterpret. This experience has only reinforced my belief that with the right team and a humble growth mindset, challenges can be transformed into opportunities for innovation and growth.

‍

Check out Shyvee's book: “Reimagined: Building Products with Generative AI”. You'll learn more about frameworks to think through which use cases should use generative AI and how companies can build moats using AI. The book features over 150 real-world examples, 30 case studies, and 20+ frameworks, “Reimagined” offers an extensive guide for integrating generative AI into product strategy and careers. Grab your copy on Amazon: https://a.co/d/btmnJfu.