Breaking through the buzz: How AI transforms unstructured feedback in CX

The advent of artificial intelligence (AI) was fast—we’ve now reached the age of AI overload. From work to your shopping list, it seems like everything is starting to involve AI. But now that we’ve arrived at AI ubiquity, you have to ask: How, exactly, is AI helping? What do you gain by using it? What are the limits?

AI assistance has been standard in the experience management (XM) industry for a while, but recent technology developments have created new opportunities to find and address customer and employee friction points faster and in a more focused way. Let’s look at what AI has to offer XM programs, where the tech has grown, and where the limitations still necessitate a human touch.

The challenge of unstructured feedback in XM

A lot of XM work is based on scoring or ranking. This strategy has a lot of different looks: stars, scores, thumbs up/down, agree/disagree spectrums, ranked lists, etc. This structured data is crucial, but it doesn’t give customers the opportunity to give feedback on what’s important to them—that’s where open-ended (or unstructured) feedback comes in.

Unstructured feedback includes anything that isn't a score or rank, and can come from almost anywhere including surveys, contact centers, online ratings and reviews, emails, and even direct messages from social media.

Open comments are rich with information but are complex to analyze for the following reasons:

• They’re volume-heavy: With feedback pouring in from across the customer journey the sheer scale is overwhelming.
• They’re unstructured: Comments don’t follow a standardized format, making it hard to sort or categorize them at scale.
• They’re highly nuanced: Language is full of context, subtlety, and emotion, all of which are difficult to decode.

Traditional systems rely on predefined categories or hierarchies to sort comments, which can limit accuracy, depth, and adaptability. Over time, customer preferences, priorities, behaviors, and language will also evolve, and static systems can’t keep up without help.

Why now is the time to embrace AI

The XM market is ripe for an AI-development leap for three key reasons:

1. Data is growing exponentially: AI offers the scalability to keep up as customers are leaving more feedback across diverse channels than ever before.
2. Competition is fierce: Exceptional CX drives loyalty and revenue, but standing out requires proactive improvement at a greater pace than ever.
3. The tech is ready: Advances in natural language processing (NLP) make it possible to accurately analyze language, tone, and intent; and today’s large language models (LLMs) can read, understand, and summarize language with exceptional accuracy.

AI-powered solutions are no longer just nice-to-have; they’re essential for businesses committed to understanding and acting on customer feedback at scale.

How AI simplifies and enhances open-ended comment analysis

AI for comment analysis isn’t just about automating what humans already do—it’s about rethinking the approach entirely. Here’s how an AI model works behind the scenes to deliver accurate, detailed insights faster:

Identifying feedback without a pre-built map

Unlike traditional tools that depend on pre-created lists of categories or keywords, a modern AI model dynamically identifies entities in customer feedback. An entity could be a product feature, a location, a staff member, or even an abstract concept like “value for money.”

For example, if a customer says, “The app was intuitive, but the checkout process took forever,” AI recognizes two key entities—app usability and checkout process. It doesn’t need a preset rulebook to do this, which means it adapts to feedback as it evolves.

Understanding relationships and context

AI doesn’t just pick out nouns—it understands how they’re connected. In the same example, the AI knows “intuitive” is a positive sentiment tied to the app, while “took forever” reflects frustration with checkout.

This contextual awareness helps surface more precise insights, so you’re not just counting mentions of a word like “app,” but truly understanding what customers are saying about it.

Adapting to new patterns and language

Customer language isn’t static, and neither is AI. As customers adopt new ways of describing their experiences (like “ghost kitchen” or “buy online, pick up in store”), the model learns and adjusts. This adaptability ensures you’re always capturing relevant insights—even when feedback shifts in unexpected ways.

AI is going deeper for XM as it gets bigger

AI has made big jumps in capability recently, most notably being able to extract significantly more color and insight from comments than was possible just 18 months ago. Traditional models would assign categories and sentiment to comments but wouldn’t be able to tell you at scale the background for why customers had the reactions they did or how to take action on it.

In contrast, newer AI models extract significantly more insight by identifying deeper, more specific categories and deducing customers’ intentions. So, what historically would have been an output of “negative sentiment towards (product) availability” becomes “negative sentiment about the availability of chicken nugget sauce options with an intention to complain.”

Industry-specific examples: How AI-powered comment analysis drives impact

AI isn’t just a technical innovation—it’s a practical tool that solves real-world challenges across industries. These are a few examples of how AI-powered comment analysis can make a difference in restaurants, retail, and service-based industries.

Restaurants: refining the guest experience

Restaurants thrive on guest feedback, but open-text survey responses can be overwhelming to process.

Example: Through AI-driven analysis, a regional restaurant chain finds a recurring theme—customers love the food quality but mention issues with long waits for orders during weekends. The AI also identifies a location-specific trend: one outlet gets frequent mentions of cleanliness concerns. Armed with these insights, the restaurant improves kitchen processes during peak hours and retrains staff at the flagged location, resulting in higher satisfaction scores and repeat visits.

Retail: enhancing the customer experience in stores and online

Retailers juggle feedback from in-store shoppers, online reviews, and customer service interactions, all of which reveal opportunities to improve the overall experience.

Example: A clothing retailer analyzes open-text feedback from customers across multiple stores. The AI flags recurring mentions of “long fitting room lines” and “difficulty finding staff assistance” in urban locations. These insights lead the retailer to optimize staffing schedules during peak hours and pilot self-service checkout options for shoppers buying without trying items. These changes improve customer flow and reduce wait times, creating a more seamless in-store experience.

Services: identifying and resolving operational bottlenecks

Service businesses, like spas, gyms, and salons, rely on feedback to constantly improve the customer journey and ensure customers are coming back and renewing memberships.

Example: A national gym chain processes customer comments from post-visit surveys using AI-enabled tools. It discovers that members at suburban locations frequently mention equipment availability during early mornings, while urban members highlight cleanliness concerns. These findings enable the chain to adjust cleaning schedules and increase equipment access during peak hours, boosting member retention.

Where AI shines, and where it doesn’t

AI is a powerful tool, but like any tool, it has strengths and limitations. Here’s an honest breakdown:

What AI does exceptionally well

• Scale: It can process thousands (or millions) of comments faster and more consistently than any human team.
• Pattern recognition: AI excels at finding themes, trends, and anomalies that might go unnoticed.
• Speed: It delivers insights in real-time or near real-time, allowing businesses to act quickly on feedback.
• Impartiality: Unlike humans, AI won't introduce cognitive biases—like groupthink or confirmation biases—into analysis.

Where AI struggles, and how we address it

AI is a powerful tool for analyzing customer feedback, but it’s not without its limitations. Understanding where AI excels—and where human expertise still plays a critical role—can help you make the most of this technology. Here are some common challenges and how our approach addresses them.

• Understanding deep context: AI isn’t great at decoding sarcasm or understanding highly complex narratives. However, it’s improving rapidly, and the newest models incorporate advanced natural language processing (NLP) techniques to handle nuance better than ever.
• Human-like empathy: AI can surface insights, but it can’t replace the empathetic touch of a human CX professional. We see AI as an enhancement, not a replacement, for thoughtful decision-making.
• Handling ambiguity: Comments that are vague or contradictory (“I loved the food but hated everything else.”) can still present challenges. That’s where human review and interpretation remain invaluable.

Why it matters: turning feedback into action

Understanding what AI is doing behind the scenes makes it easier to understand and trust the insights it delivers. Effective application of AI in your XM program will help you gain clear priorities, adaptability, and confidence to take action based on your unstructured customer feedback in a way not previously possible.

What’s next: Bringing smarter feedback tools to CX

Soon, SMG is excited to launch our AI-enabled Comment Explorer—a tool designed to provide the fastest path from open-ended feedback to actionable insights. By dynamically identifying topics, analyzing sentiment, and uncovering patterns from across the customer journey in a unified overview, it will help brands move from data to decisions faster and more effectively.

Our product roadmap includes other exciting developments as we continue to integrate this next-generation technology in 2025. In the meantime, we’re here to answer your questions about building an XM program that combines the power of human intelligence with the utility of AI to create experiences that build loyalty and improve financial performance.