Lending · Demand validation · Persona reframe

The persona reframe that reshaped Citi's lending roadmap.

Demand test and design tracks repackaging Citi's FlexPay installment-lending product for a different audience. The persona reframe (Transactor → Often Revolver) reshaped the lending product hypothesis on the strength of the test data.

Challenge.

Citi already had FlexPay in market. It let eligible cardholders convert one purchase of $100 or more into a fixed-payment installment plan. It launched on iOS in Q2 2019 with $1MM in new loans booked in April climbing to $24MM per month by November. But the product was underperforming relative to ambition. It had been built first as a Transactor incentive: give people who pay their balance off in full each month a discount to revolve a single large purchase. The Transactor segment had not adopted. The D10X Lending Pod's hypothesis was that the same product, repositioned and retargeted at Revolvers (cardholders who carry a balance), would land. The decision Citi brought us was: which Revolver segment, which messaging hook, which trigger moment, and which channel produce decision-grade demand for a FlexPay-class installment offering.

Approach.

• Two parallel workstreams: Lending Demand Tests (paid-acquisition demand validation under an unbranded "SpendFriend" front) and Lending Design (UX prototyping of the post-purchase Citi-app experience that converted the demand into a working product).
• Ad copy as the segmentation variable. Sprint 1 ran one Facebook audience with four creatives differentiated by emotional hook (Freedom, Joy, Confidence, Neutral). Cheaper, sharper, and faster than building parallel audiences with parallel creatives.
• Trigger-moment audiences over pure demographics. Job Transition, Recently Moved, Expecting a Baby outperformed pure demographic Gen Pop targeting in subsequent sprints.
• Fee-vs-APR framing tested directly. Whether Revolvers respond more to the rate or to the fixed monthly payment was a load-bearing assumption the test design isolated explicitly.

Outcome.

• Persona reframe, Transactor → Often Revolver carried into the FlexPay product hypothesis on the strength of the demand-test data.
• Balance Bifurcation surfaced as the precondition design change for the post-purchase experience, picked up by Lending Design.
• Reshape into Citi's lending roadmap, the engagement output went directly into the FlexPay-to-SpendFriend reposition planning rather than producing a recommendation deck.

Methodology lessons.

• Ad copy as the segmentation variable is a high-leverage shortcut for any audience-and-message question.
• Persona-converter delta is decision-grade signal. When Seed 1 produces a persona and Seed 2 produces an actual paid-acquisition converter pool that is different, the delta itself is the finding.
• Trigger-moment audiences outperform demographic ones for liquidity-smoothing products.