For twenty years, the honest answer to “do we understand our customers” has been: some of them, based on whoever complained loudest or filled out the survey. Generative AI has changed what’s technically possible here faster than almost any other CX capability, and that’s exactly why it’s worth a hard look at where your organisation actually stands.

The dimension is Understanding: whether customer knowledge lives in a shared, evidenced view (personas, segments, journey maps) used across the business, or whether it lives in individual heads and gets reconstructed from memory every quarter. The Key KPI is sentiment measured and segmented by persona, not blended into one company-wide NPS number that hides more than it reveals.

Before AI, building a genuinely data-led view of the customer base was expensive: research teams, structured coding of open-text feedback, journey mapping workshops that took months and went stale within a quarter. That cost is exactly why most organisations settled for Gut-feel or Partial evidence: a handful of interviews, a QBR anecdote, a strong opinion from the head of sales.

AI removes the cost excuse. Large language models can now synthesize thousands of support tickets, reviews, call transcripts, and survey responses into structured, segmented insight in hours, not quarters. The technical barrier to reaching “Managed” or “Optimised,” continuous, real-time understanding that shapes decisions across the business, has effectively collapsed.

What hasn’t changed

The floor is still governance and coverage, not tooling. AI applied to a narrow, biased sample of feedback just produces a more confident-sounding version of the same blind spot. The organisations getting real value are the ones that first widened what percentage of their customer base is actually represented in the data, then applied AI to synthesize it, not the other way around. Ask your team one number: what percentage of the customer base is covered by structured feedback in the last twelve months? If it’s under 20%, faster synthesis of that 20% won’t get you to real understanding.

The growth case

Poor experience is one of the most common reasons customers leave even when the product works, and you cannot fix what you can’t see segmented. Blended NPS numbers hide the accounts quietly disengaging behind the ones who are happy and vocal. Segmented, AI-augmented understanding is what lets growth teams target retention risk and expansion opportunity differently by persona instead of running one generic playbook for every account.

Understanding is one of 12 dimensions scored in the NextCentric Customer-led Maturity Assessment, heavily weighted as a Foundational dimension, meaning weakness here drags down everything built on top of it. If your last honest answer to “how well do we know our customers” was a shrug, that’s worth an honest evaluation before investing further in AI tooling to act on data you don’t yet have. The CSX Methodology takes that score and builds the workshop plan, scope, charter, target operating model, and ROI case, to close it.

You can take the free maturity assessment here