Delegated spending at scale.

Large-scale consumer platforms often need to support situations where one person pays on behalf of another. Think of parents supporting house managers, employers enabling field purchases or families coordinating day-to-day expenses. While this behaviour is already common, most financial systems term it as an exception, forcing people into workarounds that introduce friction, social exposure and anxiety at the point of payment.

As the UX design lead on the delegated spending system built on top of M-PESA rails, I was responsible for framing the problem, defining the system model and guiding cross-functional decisions to balance autonomy, safety and dignity at scale. Instead of focusing on transactions or controls in isolation, the efforts centered on a single outcome: ensuring that beneficiaries could complete payments with the same confidence and social ease as a primary account holder, neither transferring ownership nor eroding trust.

This case study focuses on how delegated agency was modeled as a system, including permissions, boundaries, states and trade-offs, to make shared spending feel normal in everyday, public payment moments.

People often delegate spending in everyday situations: parents enabling household purchases, employers supporting on-site work or siblings helping manage expenses. In markets like East Africa, where mobile money underpins daily life, this behaviour is common and practical. Yet most financial platforms treat shared spending as an exception, forcing people into workarounds that technically function but feel fragile and uncomfortable in real-world use.

At the point of payment, failures are public and time-compressed. There is no opportunity to explain intent or recover socially, and even brief hesitation can expose dependency or power dynamics in front of others. Designing delegated spending therefore required more than enabling access. It required making shared payments feel normal, predictable and socially safe in the moments that matter most.

Delegated spending involves two distinct roles with asymmetric responsibilities and risks. Rather than serving a single user, the system must work for people who think about money, control, and failure very differently, especially in public payment moments where confidence or hesitation is immediately visible. It was essential to design a system that could work without friction or misunderstanding.

The principles derived were not aspirational values but rather direct responses to real constraints. In a system where payment failures are visible and difficult to recover from in the moment, poor UX decisions can quickly surface as hesitation, embarrassment or loss of trust with consequences that extend beyond the transaction into everyday spending.

Several realities defined the design space from the outset. Delegated spending needed to function across uneven power dynamics, informal relationships, and diverse contexts without training, contracts, or constant oversight. The system also had to operate at national scale, where predictability and clear accountability matter more than supporting every possible edge case.

These constraints shaped the design principles that followed. Designing for only one mental model proved insufficient: prioritizing sponsor control introduced friction and awkwardness for beneficiaries, while prioritizing beneficiary flow eroded sponsor trust and predictability. The system therefore needed to enforce boundaries invisibly, thereby protecting confidence and dignity at the moment of use, while preserving accountability and control behind the scenes.

Rather than transferring funds or sharing credentials, we designed the system around a delegated agency model. Ownership of money, authority to define boundaries, and execution of payments were treated as separate verticals, each with clear responsibility. This separation made it possible to extend spending access without exposing sponsors to surprises or beneficiaries to hesitation at the moment of payment.

Not every legitimate edge case was supported. In a system involving real money and social exposure, flexibility often introduces ambiguity at the moment of use. Trade-offs were therefore made intentionally to preserve predictable behaviour, clear accountability, and confidence at checkout, even when this meant accepting friction in less common scenarios.

Designing delegated spending required close collaboration across product, engineering, and risk teams, each with different incentives and concerns. Product teams pushed for broader scenario coverage, while risk and compliance emphasized the dangers of ambiguity and misuse at scale. These tensions surfaced early and repeatedly as the system took shape.

I influenced alignment by shifting discussions away from individual features toward system behaviour and responsibility. By articulating mental model mismatches, failure modes, and accountability flows, I helped the team see how certain forms of flexibility would surface as public uncertainty or misplaced responsibility at checkout. This reframing allowed teams to agree on clear boundaries without framing constraints as compromises.

As a result, the group aligned on a boundary-first system model that balanced usability with predictability and safety. Design acted as the connective tissue between user experience, technical feasibility, and organizational risk tolerance, enabling decisions to scale consistently as the product moved into UAT.