03 · HelloFresh · 2026

Customer Care Compensation

Unifying how customer compensation is decided, explained, and delivered.

CurrentThis project is in development. The case study documents the work and decisions to date.

The current stateOne customer problem. Three separate ways to decide what happens next.
Customer contextWhat happened, who was affected, and what has already been offered?
01 · Customer initiatedSelf-service

Rules embedded in a separate flow

02 · Individual caseAgent resolution

Context rebuilt across tools and guidance interpreted manually

03 · Major incidentBulk compensation

Impact and amounts calculated through an operational process

What this createsDifferent tools, different decision paths, and no shared explanation of the outcome.
Role
Design Lead
Timeline
2026 · Current
Team
Customer Care · Product · Engineering
Skills
Product discovery · Systems design · Service design

The problem

The difficult part was not issuing credit. It was deciding what was fair, consistent, and explainable.

Compensation had evolved across separate products and operational processes. Agents could resolve an individual case, customers could receive an offer through self-service, and operations could compensate thousands of people after an incident. Each path carried only part of the context needed to make a confident decision.

My role

Design Lead
Discovery · Synthesis · Experience definition

My ownership

I lead the design work with product, engineering, customer-care operations, and compensation stakeholders. Together, we identify the highest-impact breakdowns and separate interface problems from policy, process, and system problems.

What we observedThree recurring breakdowns
01 · ContextAgents reconstructed the customer story.

They moved between tools to understand what happened, what the customer had already received, and which details they could trust.

02 · DecisionGuidance still required interpretation.

Policies did not always translate cleanly into a resolution, leaving experience and judgement to bridge the gap.

03 · ExecutionLarge incidents followed another process.

Operational teams calculated and issued compensation outside the individual case flow, limiting reuse and consistency.

Research → product direction

We started with error reporting. The research pointed to a bigger problem.

We shadowed first- and second-line customer-care agents as they moved from a reported issue to a resolution. The recurring problem was not simply categorising the error. Agents reconstructed customer context across tools, verified information they did not always trust, interpreted guidance, and sometimes made manual judgement calls.

Before

How might we improve error reporting?

After

How might the system determine the appropriate resolution while giving agents enough context to communicate it confidently?

Emerging system direction

Reframing compensation as a system, not a screen.

The emerging direction uses a shared policy layer as the source of truth across the different ways compensation is issued. A common decision model can connect customer context and compensation rules while allowing each channel to support its own workflow.

Design principle

The system can be complex. The agent experience should not be.

01

Understand the customer’s situation

02

Capture the relevant information

03

Receive the appropriate resolution

04

Understand enough to explain it

05

Communicate it confidently

Current · Work in progress

Where we are now.

The system direction is being developed and validated across three connected workstreams.

01

Policy configuration

Refining how compensation stakeholders configure the rules that govern compensation across channels.

02

Agent resolution

Working backwards from frequent customer scenarios to define what agents need to know, capture, and explain.

03

Bulk compensation

Mapping incident workflows and identifying where manual calculation can be replaced by the shared compensation model.

Next phase

Validate policy configuration, define the highest-value agent scenarios, and understand the upstream bulk-compensation workflow.

Next project · 04KrungThai Banking App