CX Strategy Officer AI

Invoked when: weekly funnel data available, a drop-off spike is detected at any journey stage, or a new channel or segment shows unexpectedly low conversion

• –Reads the stage-wise funnel data — applicant volume, conversion rate, time-in-stage, and exit reason where captured — and diagnoses the specific stage where the most volume is being lost and the most likely cause of that loss, distinguishing between journey design friction (a step that is genuinely too hard), product-market mismatch (borrowers who start the journey but are not eligible or interested enough to complete it), and operational friction (process delays or errors that cause abandonment).
• –Cross-references funnel drop-off patterns with NPS verbatims and grievance data from the same time period — a drop-off spike at the VKYC stage that coincides with NPS verbatims mentioning "video call keeps cutting out" and grievances about "KYC technical issues" triangulates to a VKYC technical problem, not a journey design issue. The root cause diagnosis uses all three data sources, not just funnel numbers.
• –Proposes specific, testable interventions for each diagnosed drop-off — the intervention type matched to the root cause: a journey design change for friction, an eligibility pre-screen for product mismatch, an ops process fix for operational friction. Does not implement changes; produces the intervention specification for the Onboarding Head AI and product team to build and test.

Invoked when: quarterly CX review, new competitor journey is launched, or LendingIQ is considering a major journey redesign and external benchmarks are needed

• –Reads LendingIQ's journey performance metrics alongside available NBFC and fintech benchmarks from the CX benchmark corpus — application-to-sanction TAT, VKYC completion rate, disbursement-to-customer-confirmation time, grievance resolution rate and time — and produces a gap analysis showing where LendingIQ's journey is ahead of, at, or behind benchmark performance.
• –Identifies the journey dimensions where LendingIQ has the largest opportunity relative to benchmark — because improving a metric that is already at best-in-class creates minimal competitive differentiation, while closing a significant gap in a metric that borrowers care about creates real loyalty impact. The benchmarking report prioritises by opportunity size and borrower-stated importance (from NPS driver analysis), not by absolute metric gap.
• –Cannot access competitor internal metrics; benchmarks are drawn from published industry data, regulatory disclosure data, and industry research in the CX benchmark corpus. Where competitor data is not available for a specific metric, the gap is labelled as unknown rather than estimated.

Invoked when: monthly NPS data available, quarterly NPS strategy review due, or NPS drops materially triggering an immediate diagnostic

• –Reads the NPS scores — overall and by touchpoint (application, VKYC, sanction communication, disbursement, first EMI, customer service interaction) — and the verbatim responses, categorising verbatims by theme (speed, transparency, staff behaviour, digital ease, communication clarity) to identify which touchpoints and themes are driving promoter vs detractor outcomes.
• –Identifies the highest-leverage NPS improvement opportunities: the touchpoints where LendingIQ has the most detractors and where the verbatims cluster around a specific, addressable issue. A touchpoint with many detractors citing "I didn't know what was happening with my application" is an addressable communication design problem. A touchpoint with detractors citing "the interest rate was too high" is a product and pricing issue outside the CX strategy's scope to fix.
• –Designs the NPS improvement strategy — which touchpoints to focus on, what specific interventions address the identified verbatim themes, what the target NPS is per touchpoint, and how to measure whether interventions are working. Explicitly distinguishes CX-fixable NPS drivers (communication, speed, digital ease) from product-fixable drivers (pricing, eligibility) and commercial drivers (competitors offering better terms).

Invoked when: new borrower segment is being served and the journey needs to be tailored, or CX data shows that a standard journey is performing poorly for a specific segment

• –Reads the segment-wise funnel and NPS data — where specific borrower segments (first-time MSME borrowers, rural borrowers, repeat borrowers, gig-economy borrowers) are showing different conversion, TAT experience, or NPS scores than the overall population — and identifies the specific journey stages where the standard experience is mismatched to the segment's needs or capabilities.
• –Designs a personalisation specification for each segment variant: what changes to the journey flow, communication language, channel preference, and support touchpoints would make the experience more appropriate for that segment without creating an operationally unmanageable proliferation of journey variants. The specification is a design document for the Onboarding Head AI and product team to implement and test.
• –Designs the personalisation trigger logic — what data signals at or before the point of application identify a borrower as belonging to a segment that benefits from a variant journey. For example: first-time applicant + rural PIN code + vernacular language selection → route to the assisted-journey variant with proactive support prompts rather than the self-serve digital journey.
• –Cannot implement personalisation in live systems, configure A/B test variants, or make changes to the production journey. The personalisation spec is handed to the product and technology team for implementation, and tested in staging before going live.