Mattel’s game changer: How AI is turning customer feedback into real-time product updates

Nobody likes being stuck in an elevator — including Barbie.

Last holiday season, families began noticing glitches with the elevator in their Barbie Dreamhouses. As often happens these days, they began turning to social media and the review sections of major online retailers to vent their frustration.

In a feat bigger than He-Man and more coordinated than an American Girl, the customer service team at Mattel collected and catalogued nearly every single complaint and quickly worked through their design and supply chains to fix the problem. In a matter of days, Mattel was able to achieve an almost instantaneous renovation for future Dream Homes.

The handyman that made this handiwork possible: Gemini.

“Through the power of Google, we're now able to turn the written reviews into consumer sentiment from our global consumer service area, and we can do that instantaneously and at scale,” Shaun Applegate, Mattel’s director of product quality analytics, told us in a recent interview. “We're no longer limited to individual members working, doing the grunt work, doing the tally sheets — with AI, we just have a visual and a report out of what our consumers are talking about like that.”

And this is just one example of at least a half-dozen where Mattel is deploying generative AI across their business, including marketing, forecasting, customer service, design, and e-commerce. It's all aimed at reimagining the way Mattel makes toys, and dreams, come true.

Not playing around with AI

In the toy industry, the secret to a smash hit requires more than simply a passion for design, engineering, and creative ideas. A deep understanding of how fans experience and imagine the world around them is a crucial element — especially as fandoms have evolved to include more age groups and demographics.

Since its humble beginnings in a garage in 1945, Mattel has consistently been reshaping play for children and families across the globe with iconic franchises like Barbie, Hot Wheels, and Fisher-Price. This enduring success is not only a testament to its relentless drive for innovation but also Mattel’s innate ability to capture the wonder of childhood and bring it to life.

Today, this task is proving harder than ever before, even for the masters at Mattel. How we play, what we play with, and even who’s playing is rapidly changing. Toy companies and manufacturers have long relied on traditional trend-spotting and a good gut instinct. Now they face unprecedented disruption fueled by digital technologies, fleeting viral trends, and the rise of new target demographics like “kidults” who are redefining the very concept of play.

Mattel has always brought the same rigor to designing its internal systems as it has the toys those systems create. In the digital age, data and AI have taken on an increasingly important role.

“Our focus is to turn data to insights, driving good decisions around making the best quality products on the planet,” Applegate said. “Play is the center of all we do. It’s a special place to be when you can put that product in the hands of kids and see the joy — it’s just an exciting opportunity for us to work every day to make better toys with each iteration.”

More than ever, that all-important data isn’t just numbers in a spreadsheet — it’s the key to deciphering customer behavior, nailing demand forecasts, and transforming how products are created and sold. Thanks to the expansive power of AI, Mattel is now gaining a deeper understanding of its customers at speeds that were previously impossible.

From reactive to actionable insights

The ever-growing expectations of customers means even an industry leader like Mattel has to constantly rethink the way it understands customers to better meet their needs. In particular, Mattel’s consumer feedback process had long been a bottleneck, with data analytics teams struggling to sift through enormous volumes of customer interactions across diverse channels.

“When I started three years ago, employees were writing tally sheets and taking days, weeks, months to report out information,” Applegate recalled, meaning insights often arrived too late to inform decision-making or spot emerging opportunities. “We were basically talking about things of the past that nobody was interested in. It’s not where we wanted to be.”

Furthermore, Applegate says that relying on manual processes prevented analysts from gaining a complete understanding of product performance or identifying broader issues.

“We didn’t know what we didn’t know, and we certainly missed a lot, and really what we were missing was a holistic perspective,” he said. “You didn’t know if that issue with that one individual product was actually a larger brand issue, a manufacturing issue, or a supply chain issue.”

To continue its legacy of empowering generations to reach their full potential, Mattel recognized it needed to do more than just listening after the fact — it needed to actively understand what customers need and want in the moment.

Unlocking the voice of the consumer

To improve its understanding of consumer sentiment, Mattel developed an AI-powered feedback classification system as part of its newly announced partnership with Google Cloud. The cutting-edge system, powered entirely by Google Cloud’s advanced data analytics and AI capabilities, can analyze millions of customer interactions from a diverse range of sources in a matter of seconds.

“Under the hood, we’re running everything in Google Cloud,” T.J. Allard, Mattel’s lead data scientist explained during the interview. “So, we can do from beginning to end — ingestion to AI inference — all in one fell swoop.”

At its core, the system relies on BigQuery for storing and efficiently processing its massive customer datasets, often containing tens of millions of customer feedback points from social media, online reviews, and direct customer communications. With this strong data foundation in place, the team then utilized Vertex AI and Google’s multimodal Gemini models to refine and train a sophisticated consumer feedback model that can accurately classify different types of customer feedback and interpret the underlying sentiment.

Together, this powerful combination enables Mattel to translate consumer feedback into valuable insights that would be impossible to gain through manual methods.

“We’re not limited to individual members working, doing the grunt work, doing the tally sheets,” Applegate said. “Through the power of Google, we’re able to take the voice of the consumers from data to insights on demand.”

Choosing to build the solution internally was another “aha moment,” according to Applegate.

“Realizing we could do this internally made a big impact,” he said. “With our small team, we were able to do what we wanted and get the platform we wanted, and we realized around $1 million in cost savings.”

From months to minutes

Already, the new AI-powered system has delivered significant wins, delivering a staggering 100x boost in data processing capacity and reducing analysis times from a month to a single minute.

“Our big motto is ‘From months to minutes,’ but it’s real,” Applegate said. “We were literally spending months-worth of analysis and just getting data into the place that an analyst could tally up all the sentiment — and now it’s just at our fingertips.”

In addition to the Barbie Dreamhouse work benefiting from AI, other areas Mattel has been experimenting with the new technology include sustainability, brand health, and product quality. They have begun automating the data collection, validation, and narrative report generation that goes into meeting new European Union reporting regulations. They’ve also augmented the human analysis that goes into brand quality reviews — a quarterly brand audit of sorts, based on product performance — and product development reviews — which spot potential production and quality issues.

As Mattel continues to advance its data and AI capabilities, Allard believes the new customer feedback solution has set the stage for future success. “I think where we go next with this is even deeper insights,” he said. “Truly leveraging AI modeling to understand what is the core of what our consumer is trying to communicate to us, so we can immediately act on that. Turning the volume of information into value inside the organization.”

This new AI-driven approach aligns with Mattel’s broader vision to empower teams to explore data in any way necessary.

By automating the analysis of many processes, analysts are now freed from the noise of everyday tasks, enabling them to focus on deeper research across the company’s iconic portfolio brand.

With more than a hundred brands, each with its own caveats about development and design, Mattel can now accurately assess what consumers are looking for, driving tangible product enhancements and personalized experiences that forge greater connections with its fans.

“Now, because we can do this at scale and on demand, we’re able to slice and dice this information however we want. It’s really opened the doors and frankly, the eyes of our analysts about what the possibilities are,” Applegate said. “Taking this data systematically and being able to report back to the global brand teams has been a game changer.”