Traditional funnels miss the nuance of individual buying journeys. Conviva's CEO argues for analyzing personal behavior patterns—like a "research shopper" toggling between cart and reviews—to understand user intent and boost conversion, rather than forcing users into a predefined sequence.
An agent that ignores a user's preceding on-site behavior creates a frustrating experience by forcing them to waste time re-explaining their context. To be effective, agents must be fed the user's session data to start the conversation with informed, relevant suggestions or questions.
Conviva's CEO warns against "AI washing," where companies apply AI agents to poorly structured data. An agent cannot invent insights that aren't present in the source data. A strong data computation engine is the true prerequisite for effective AI, not a cosmetic front-end.
Instead of a top-down approach where you must know what questions to ask your data, Conviva's CEO advocates for a bottoms-up methodology. Collect all behavioral data, compute patterns, and let the platform automatically surface trends and interruptions, telling you where to act.
Many product teams react to any user friction, like a bounce. Conviva's CEO argues this is inefficient. Instead, teams should differentiate between low-intent "bouncers" and high-intent users who faced a specific obstacle. Solving problems for the latter group yields a much higher ROI on engineering and design time.
While product teams design simple 6-7 step funnels, data from Conviva reveals that the average real-world e-commerce buying journey involves over 50 steps, sometimes even 150. This starkly illustrates how much user behavior and intent is lost in traditional, oversimplified analytics models.