Omer Kehat on AI-Powered Consumer Insights, Cross-Functional Collaboration & the Future of VoC at Revuze

1. What measures are in place to ensure that voice-of-customer data from various channels (e.g., reviews, social media, surveys) is effectively utilized to inform business decisions?

Brands today are inundated with Voice of Customer data from every direction. To turn this chaos into clarity, data quality must come first. It is the foundation, the building blocks, that drive informed decisions across every industry. At our core, we are focused on delivering the highest-quality insights to brands.

We started with review data, a post-purchase source making it highly credible but often challenged by duplication due to syndication. Once we perfected that, expanding our AI engine to handle social media and survey data was a natural next step.

Revuze does not stop at analysis. Our AI goes further by offering data-backed recommendations and activities that help brands move confidently from insight to action.

2. What technologies are currently employed to process and analyze large volumes of unstructured consumer feedback, and how do they integrate with your existing systems? 

Since 2011, Revuze has been investing and optimizing its own proprietary Generative AI and LLMs, well before ChatGPT took the spotlight. Over the years, we've focused on refining and optimizing our large language models to deliver unmatched precision in analyzing VoC data.

Our GenAI engine detects tone and context with remarkable accuracy. Take battery life, for example. It's important across many industries, but context matters. A one-hour battery life is unacceptable for a smartphone, yet impressive for a drone. Our AI is trained to understand these nuances based on the product category, ensuring insights are both relevant and accurate to support brands’ decision-making.

This deep expertise powers best-in-class topic extraction and sentiment analysis across more than 1,500 product categories. In addition, Revuze integrates advanced "off-the-shelf" LLMs, such as Anthropic’s Sonnet, for specific tasks like generating review summaries, further enhancing our platform's capabilities.

3. In what ways are you streamlining cross-functional collaboration between departments (e.g., marketing, product development, customer service) to act on AI-driven recommendations?

One of the core challenges many companies face is siloed data, where different teams rely on separate sources. This often results in inconsistent insights and misaligned decisions. ActionHub addresses this by bringing all teams together, marketing, product, eCommerce, and consumer insights, on a shared foundation. While each team has a customized hub tailored to their needs, everyone works from the same unified data set. This includes review and rating data, social media, and survey responses, all accessible in one place.

The platform combines quantitative and qualitative data in a visual format, making it easier to uncover not just what consumers are saying, but why they are saying it. Regardless of role, users can explore the drivers behind consumer behavior in a way that is both intuitive and consistent.

To encourage collaboration, we have developed cross-functional use cases that help teams align on opportunities and next steps. For instance, product teams can use insights from the ProductHub to identify purchase motivators and align them with product strengths, guiding marketing on which messages are most likely to resonate. Similarly, marketing can collaborate with eCommerce teams to refine product detail page (PDP) content based on consumer feedback, improving online performance.

We have also prioritized integration with enterprise tools such as Microsoft Copilot and Claude. This allows organizations to combine VoC insights with internal sales data, enabling a more holistic view of the customer and supporting advanced querying.

4. How does your organization track and analyze the effectiveness of AI-generated insights in improving product features, marketing campaigns, and eCommerce performance?

At the core of the Revuze platform is a recommendation engine powered by proprietary large language models, which is also supported by external LLMs for tasks such as summarization and validation. The system is designed to turn VoC data into practical, tailored recommendations across a wide range of use cases.

In the MarketingHub, this includes support for content creation, whether it’s creating social media posts that resonate, enriching product detail pages, or creating influencer kits and data-backed video scripts - all based on the VoC data. On the product side, the ProductHub offers data-driven suggestions, ranging from addressing specific product pain points to effective innovation planning.

A key focus of the platform is helping teams explore new ideas. For example, product managers can begin with addressing unmet needs that are based on consumer review data, and the AI engine surfaces relevant trends from social media, drawing inspiration from entirely different industries. A concept in footwear might emerge from patterns identified in automotive or electronics. The data sets are being used in different ways, ways that always leverage their strengths.

These examples illustrate how the platform is built to support cross-functional teams in turning consumer feedback into actionable insights and creative exploration.

5. How is your organization preparing for emerging trends in consumer behavior analysis, such as the use of large language models (LLMs) and advanced AI in deriving insights?

Our organization is closely engaged in the advancement of AI and LLMs, with a focus on how they enhance consumer behavior analysis. We see LLMs as tools that complement traditional research methods by automating tasks like sentiment analysis, theme detection, and text summarization. This speeds up the research process and allows teams to access deeper insights with agility.

We use a combination of proprietary and third-party models, applying each where it adds the most value. Internal models help with contextual understanding in specific product categories, while external models support tasks like content summarization.

Our approach remains flexible and research-driven, allowing us to adapt quickly to new technologies while maintaining accuracy, relevance, and responsible AI use. As consumer data sources evolve, we are committed to evolving with them to support timely, data-informed decision-making.

6. How are you ensuring that your approach to consumer insights remains adaptable to evolving technologies and competitive pressures?

We stay informed about technological developments and evolving industry trends, continuously exploring new tools and methodologies. By maintaining an open and collaborative approach with partners and actively testing emerging technologies, we strive to remain adaptable to a wide range of research needs.

Our platform is designed to support both quantitative and qualitative data analysis and to accommodate various research approaches and needs. This flexibility allows us to work effectively with organizations across different sectors, helping them access and interpret the insights most relevant to their goals.