Chloe Maraina is a specialist in transforming complex data into compelling visual narratives. As a Business Intelligence expert with a deep focus on data science, she bridges the gap between raw information and strategic integration. Her work focuses on how global enterprises can leverage emerging technologies to simplify the customer journey and turn massive datasets into actionable insights for marketing and digital experience teams.
Key themes of our discussion include the unification of fragmented regional data, the superiority of agentic AI over traditional data lakes for analyzing unstructured media, and the shift toward “concierge” digital shopping experiences. We also explore the necessity of organizational guardrails to prevent fragmented AI development and the strategic importance of owning proprietary on-device agents to ensure security and long-term competency.
Large enterprises often manage dozens of separate regional operations and internal data stores. How do you unify structured and unstructured data to simplify the customer journey, and what specific steps ensure these insights reach local marketing teams effectively?
The primary challenge for a global organization, like one operating across 35 countries and China, is that data often becomes incredibly disaggregated. To simplify the customer journey, we first have to “sew together” information from disparate marketing campaigns and internal data stores. The integration process begins by aggregating standard website metrics—like conversion rates and click-through rates—and then layering in unstructured data from media and localized campaigns. This creates a unified view that allows local marketing teams to explore new campaign ideas without waiting for a central analyst to process the request. By putting these tools directly into the hands of the front-line teams, they can identify the “little insights” that scale tremendously across the profit and loss statement.
Traditional data lakes often struggle to integrate unstructured information from media and website traffic. Why is agentic AI a more effective tool for analyzing media effectiveness alongside conversion rates, and what are the primary hurdles when moving beyond standard database queries?
While a data lake is a common solution, it often fails to capture the “thorny brambles” of unstructured data, such as the specific content working in media versus what is driving actual traffic. Agentic AI is significantly more effective because it can ingest and analyze these unstructured elements simultaneously with structured data, greatly expediting the feedback loop. The primary hurdle when moving beyond standard queries is the sheer volume of data; it simply takes too long for human analysts to manually marry media spend with website attach rates. By using an agent, we can move beyond just asking “what happened” to understanding the “why” behind media effectiveness, allowing for a more nuanced analysis of how different content types influence the final conversion.
Most small business buyers begin their journey online, yet expansive product catalogs can be overwhelming. How can AI agents act as a “concierge” to narrow search results, and what parallels can be drawn between this digital experience and a personalized, high-end retail interaction?
Since roughly 70% to 80% of small business buyers start their journey online, the digital interface must be frictionless. We view the AI agent as a digital sommelier; just as a sommelier doesn’t expect you to know every bottle on a massive wine list, our agent shouldn’t expect a buyer to navigate a dense technology catalog alone. The agent asks a few targeted questions—much like asking if you prefer red or white or what your budget is—and then provides a “good-better-best” recommendation. This concierge experience transforms a confusing search into a guided conversation, allowing the customer to find exactly what they need without feeling the weight of a massive product portfolio.
When multiple departments develop independent AI use cases, organizational efforts can easily become fragmented and inconsistent. What guardrails are necessary to ensure teams work from the same data, and how do you orchestrate these activities to maintain a cohesive business strategy?
Fragmentation is one of the hardest issues to manage because excitement leads to many “boats on the water,” often rowing in different directions. To maintain a cohesive strategy, we must implement strict guardrails and guidelines that ensure every team is working off the same foundational data. This requires a formal governance element focused on knowledge management and the prioritization of specific use cases. Orchestration is the key word here; it isn’t just about control, but about managing the business activities of agentic solutions so that the probabilistic nature of AI produces consistent, reliable results across the entire organization.
Measuring the benefit of custom on-device agents involves both long-term strategic goals and immediate KPIs. Beyond media spend efficiency, what specific metrics indicate a successful deployment, and why is it beneficial for a company to own the agent rather than outsourcing it?
Success is measured through a mix of immediate efficiency gains and long-term strategic positioning. While we look at basis points of efficiency in media spend, we also focus on the “decision matrix” of owning the core competency. By building and owning an on-device agent like Qira, a company ensures that data remains secure and that the user experience is natively integrated into the hardware. The strategic trade-off is that while development is a significant investment, owning the agent allows the company to control the entire ecosystem and develop a deep, internal expertise that third-party solutions cannot replicate.
What is your forecast for agentic AI?
I believe we are rapidly approaching an inflection point where agentic AI will become the absolute norm, effectively replacing the traditional search behaviors we see on platforms like Google today. As users move toward a conversational interface with agents like ChatGPT, Claude, and Gemini, the expectation for a “concierge” experience will become standard for both general consumers and B2B buyers. In the very near future, the idea of navigating a static website will feel outdated; instead, every digital interaction will be a personalized, real-time dialogue where the technology anticipates the user’s needs and provides immediate, curated solutions.
