Loyalty economics · Experience design · Voice of the Customer
I turn customer insight into operating decisions.
I own loyalty economics end to end at a regional fintech — the programme, its cost centre, and the monthly transacting user target — with the customer research underneath to know why the mechanics work.
A decade across Malaysia, Singapore and Indonesia. I work in build and transformation moments, where the problem is not yet defined and the outcome is shaped with the executive team, not handed down.
The shift I work on
From research that gets filed, to research that changes what gets built.
Where most organisations are
- Customer signal stuck in decks no one revisits.
- Findings softened to keep stakeholders comfortable.
- Research commissioned after the decision is made.
- Insight treated as a reporting function, not a strategic input.
- "Just talk to users" repeated as advice, with no system behind it.
Where the work lands
- Findings that survive scrutiny because the methodology does.
- Teams that know what to ask before they brief.
- A Voice of Customer infrastructure leadership actually uses.
- Insight that changes roadmaps, not slide decks.
- A research function the rest of the organisation takes seriously.
The method underneath is problem-space research, built on Indi Young's framework — the structural reasoning behind customer behaviour rather than the behaviour itself. It holds up under scrutiny because it goes to the conditions and mental models that shape what people do. The discipline it demands is resisting surface explanations: the significant finding is rarely in the first answer a customer gives.
What I am accountable for
Loyalty points redemption rate, sustained — the measure that says a programme is used, not merely accrued
Loyalty cost as a share of eligible GTV, held inside a tenth-of-a-percent band
2.8B Boost loyalty points 'burned' (i.e. engagement) under my ownership, Sep 2025 - July 2026
Lift in voucher redemption demand from the Flash Deal mechanics I introduced
Case study · Boost, Axiata Group · 2025–2026
Holding loyalty cost flat while redemption stayed at 80%.
The problem on the table
The usual lever for a loyalty programme running hot is to restrict it — devalue the currency, cap the catalogue, slow the earn. That protects the cost line and quietly kills the behaviour the programme exists to buy. I was given the programme, the cost centre and the MTU target at the same time, which meant there was nowhere to hide a trade-off.
The decision I made
Manage the burn mix, not the burn volume. Move demand toward lower-cost formats members actually prefer, and widen where loyalty points are useful so the same spend buys more perceived value. Members kept full access; the economics changed underneath them.
Shift the mix, not the access
Pick and Win and similar formats absorbed redemption demand at a materially lower unit cost than catalogue rewards — without removing a single option from members.
Make someone else fund the reward
Scaled Partner Wallet into a merchant-funded ecosystem covering groceries, petrol, utilities and telco. Redemption utility moved into everyday spend; 69% of surveyed loyal users adopted it.
Target instead of broadcast
I define the targeting logic and redemption mechanics for personalised loyalty campaigns; CRM runs the sends. 8× click-through and a 58% lift in email engagement against audiences 20× smaller than mass sends.
Rebuild the demand moment
Revived BoostUP Day: daily transactions up 154% and daily net revenue up 256% on campaign days. Flash Deal mechanics lifted voucher redemption demand 20×.
Let members pick the cheaper option
FlexiRedeem gave members partial-value redemption instead of all-or-nothing thresholds. 70.9M loyalty points burned through it — volume moving into the format that costs the business least, by member choice rather than restriction.
Know which mechanic to build
None of the above is guesswork. Voice of Customer infrastructure across four platforms — Zendesk, social listening, in-app surveys, NPS — tells me which mechanics earn redemption before we spend on building them.
| Measure | Result | Why it matters to the P&L |
|---|---|---|
| Loyalty cost / eligible GTV | Held inside a 0.1x% band | A predictable cost line the CFO can plan against |
| Loyalty points redemption rate | 80%, sustained | Liability is converting to behaviour, not sitting on the balance sheet |
| FlexiRedeem burn | 70.9M loyalty points | Members choosing the format that costs the business less, unprompted |
| BoostUP Day, campaign days | +154% transactions · +256% net revenue | Loyalty mechanics moving top-line, not just engagement |
| Personalised loyalty campaigns | 8× CTR · +58% email engagement | Same outcome from a fifth of the audience and the send cost |
The finding was never the deliverable. The mechanic that changed the number was.
The research record · Boost, Traveloka, MABECS
Research is only credible when someone changes a decision because of it.
The loyalty numbers above are downstream of this. What I actually build is the capability that makes customer signal usable — and the evidence it worked is that other people's decisions changed.
Moved the cadence · 5 years, 4 platforms
From quarterly decks to weekly decisions
Five years of problem-space research at Boost, and Voice of Customer infrastructure built from scratch across four platforms, changed when customer signal arrives — from a quarterly readout to a weekly input product teams plan against. The PayFlex NPS result is the one the Chief Commercial and Technology Officer cites; the exec forums it now runs through are the part that outlasts me.
Changed the brief
Teams that know what to ask
I raised research standards across product and design teams that were not mine to manage, and coach senior researchers running their own work. The measure is upstream: better questions arriving in briefs, before anyone commissions a study.
Moved a win rate
Research that won the client
At MABECS a research-led rebuild changed how prospective clients chose between them and their competitors — established in client interviews after launch, not inferred from traffic.
"We have changed the way we operate our business almost entirely." — Executive Director, MABECS
What the executives I worked for say
You have done magic at Boost and really championed the voice of the customer in all our forums. Our PayFlex Customer NPS is a testament to that.
We have changed the way we operate our business almost entirely. Anecdotal evidence from clients suggests that after visiting our website they are more likely to choose to work with us over our competitors. The thoughtful research process is creating content that is user led and incredibly rich.
Pei Ling acts as someone who provides helicopter view to the team. She makes sure that we always look back to the core strategy by asking the right questions.
Her problem solving is one of the most methodical and data driven I have encountered. She genuinely cares for every client she works with.
She helps me identify my blind spots and gave me time to reflect on things I would not be able to do on my own — professional perspectives I apply in day-to-day decision making.
It was a joy to collaborate with her on projects. As a conscientious project manager, she never left people behind. She responsibly took everyone to keep up with her plans.
The mentorship has become an important part of my personal-professional support network. Pei Ling is an important sounding board when it comes to discussing designers' well-being, recruitment and performance review.
Leadership
I build the function, then the function outlasts me.
1 → 18
Founding hire for Traveloka's expansion out of Jakarta into Malaysia and Singapore. Grew the function to eighteen people by 2017, across copywriters, UX writers, research, visual design and editorial in three markets.
5
A team spanning Voice of Customer, Touchpoint Investigation, Journey Management and Experience Transformation — built so customer signal reaches loyalty and product decisions directly, without a translation layer.
4
Multichannel Voice of Customer infrastructure built from scratch across Boost Life, Boost Biz, Boost Digital Bank and Boost Credit — Zendesk, social listening, in-app surveys and NPS.
VP
One of the senior design leads I mentored is now a VP of Design. I coach senior researchers and set research standards for product and design teams running their own work.
How I work with an executive team
I take the undefined problem, not the specified one. I bring the commercial frame and the customer evidence in the same conversation, so the decision gets made in the room rather than deferred to more research. I am comfortable owning the number that results.
What I build myself
I ship working agents and applications without engineering support — Claude Code, Cowork, Copilot Studio, Lovable. This site is one of them. It means I can prototype an operating change in days and put it in front of the people who have to approve it.
Where I came from
I started in language — linguistics, journalism, publishing and translation. It left me with a particular way of reading organisations: what they say, what they mean, and where the two diverge. How something is framed determines whether it changes anything.
Published work
"You really pause when someone asks you a question like that."
I write Between Silences — essays on research methodology, organisational dynamics, and the leadership conditions that make rigorous work possible. It is where I think in public about the part of the job no one teaches: making insight survive the room it lands in.
Read Between Silences →If you are shaping the next transformation, I would like to be in that conversation.
Open to conversations about research and design leadership roles, advisory and consulting work, and speaking.
Specific problems make for better conversations than general interest — more customer signal than you know how to act on, a metric moving the wrong way, or a research function that needs to be taken seriously.
Pei Ling Chin · Kuala Lumpur, Malaysia · peiling.work