Four Levels Of Customer Understanding

Hannah Shamji outlines a hierarchy of customer insight that starts with the easiest-to‑collect but most superficial “what they say” (surveys, CRM entries) and ends with the hardest‑to‑capture “why they do it” (root motivations uncovered through repeated, trust‑based interviews). She backs each tier with concrete examples: the limited value of Net Promoter Score, the distortion introduced by “speak‑aloud” protocols, and the ambiguity of verbal probability terms demonstrated in Thomas D’hooge’s Dutch study. By recommending direct observation of clicks, scrolls, and pauses, and by introducing tools such as Geoffrey Roberts’s Emotion Wheel, the piece supplies a practical roadmap for teams that want to replace guesswork with evidence‑based design decisions.

The framework arrives at a moment when UX teams are shifting from purely qualitative feedback loops to blended approaches that fuse analytics, heat‑maps, and ethnographic methods. Companies like FullStory and Hotjar have popularized session‑replay and behavioral metrics, while design consultancies such as Nielsen Norman Group continue to champion deep user interviews. Shamji’s model bridges these worlds, echoing Erika Hall’s warning against asking “burning questions” and reinforcing the industry‑wide push to validate intuition with observable behavior. The critique of NPS also mirrors a broader reassessment of single‑metric loyalty scores, as firms adopt more nuanced sentiment gauges and longitudinal studies to predict churn.

If organizations adopt the four‑level schema without allocating sufficient time for the “why” tier, they risk superficial fixes that mask systemic issues. Over‑reliance on click data can still miss emotional friction points, while excessive interview cycles may stall product velocity. Teams should therefore prioritize a balanced research cadence: start with broad behavioral logging, layer in targeted sentiment probes, and reserve deep‑dive interviews for high‑impact user segments. Watching how emerging tools integrate emotion‑tagging directly into session recordings will be key to scaling this approach without inflating research budgets.