Design as a Force for Good
We built the disruption. Who paid for it?
We love products that disrupt, but do we ever stop to ask, who is that disruption for and is that ever for something good?
I believe design is a force for good.
I had a conversation recently that stuck with me. Take Gojek or any gig-based delivery services. From the get-go, you can say it helps create “jobs” or at least “living” for its service providers. Drivers earn a living, customers get things done relatively quickly. For developing markets, or for governments who don’t care, this is such a great idea. More people can get on their earning journey helping their livelihood fast.
But on the other side, there is no employment contract. No insurance. No protection from abuse.
In 2022, researchers documented that between 2015 and mid-2022, at least 89 motorcycle drivers from app-based platforms in Indonesia died in road accidents. The same year, two Gojek drivers in Indonesia separately took their own lives. Both were reportedly under severe financial pressure. One was found near a railway crossing in Bekasi, still holding the smartphone he used to take orders. The other jumped from the Suramadu Bridge, wearing his Gojek jacket.
And this happened after Gojek and Grab had been quietly cutting driver incentives since 2017, in some cases by as much as 50%, while simultaneously raising performance targets. Drivers had to work 10 to 15 hours a day just to chase the same bonuses they used to earn faster. When drivers protested in front of Gojek’s headquarters, the company responded by saying they were involved in “routine discussions.” The incentives were cut anyway.
People might despise this but they will unconsciously rely on these services for most trivial cravings. I was also guilty of pushing through a few deliveries no matter the distance, taking these for granted, thinking I deserve these forever.
There’s actually a name for this tension in tech policy: The Collingridge Dilemma.
The idea is simple: when a technology is new and easy to change, we can’t yet see its problems. But by the time the problems are obvious, the technology is already everywhere, deeply embedded, and very hard to fix.
But here’s the thing that really gets me. In many of these cases, the companies were not blind to the problems. They just chose profit anyway.
In 2022, 124,000 internal Uber documents were leaked to the press. The Uber Files, as they became known, revealed that while Uber was publicly positioning itself as a way to empower drivers, internally it knew its model relied on weak labour laws. It lobbied governments aggressively to keep those laws weak. In one reported instance, an executive even suggested the company could use violence against protesting drivers as a PR opportunity. This wasn’t naivety. This was a calculated choice.
Uber spent years fighting driver reclassification in court across multiple countries, while simultaneously spending over $200 million on a California ballot measure to carve itself out of a law that would have given drivers basic rights. They won that vote. The drivers lost.
The same pattern is playing out across the the world, just faster and with fewer checks. In Malaysia, Grab drivers went on strike in 2022, and again in a “Grab Blackout” in 2024, protesting opaque pay systems and unsafe conditions. In Kenya, Nigeria and South Africa, gig workers are organising and pushing back. Gig platform job listings in Sub-Saharan Africa grew by 130% recently, one of the fastest rates in the world. The infrastructure of exploitation is scaling before any protection does.
By the time we realised ride-hailing apps were hollowing out labour protections, millions of drivers were already dependent on them. Changing the model now is expensive, politically messy, and slow.
So what’s the alternative? One answer is the Precautionary Principle: don’t deploy until you can prove it won’t cause harm. But that arguably kills innovation before it starts. No startup would operate on this principle.
But maybe the real answer is not to wait for regulators to catch up. Maybe it starts with the people building these products asking harder questions earlier.
What happens to a driver if they get injured and can’t work for a month? Can the incentive structure I’m designing push someone to work dangerously long hours? If I add a new efficiency feature, who bears the cost of that efficiency?
Singapore’s Platform Worker Act, rolling out in 2025, is at least trying to build protections in without dismantling the flexibility model. It’s not perfect. But it proves the question was possible to ask earlier, and that someone eventually did.
Growth and ethics don’t have to be enemies, I guess. But that only works if the people with the power to design these systems stop treating ethics as someone else’s department.
Because here’s the thing. Ethics doesn’t live in the legal team or the policy floor. It lives in the product decisions made way earlier than that. It lives in the designer who decides whether to show a driver their weekly earnings clearly or bury it behind three screens. It lives in the PM who sets the acceptance rate threshold that determines whether a driver gets deactivated. It lives in the engineer who builds the algorithm that decides who gets orders and who doesn’t.
These are not neutral decisions. They never were.
I can’t claim I’ve always gotten this right. But I think about it more now. When I’m working on a product, I try to ask: who is the most vulnerable person in this system, and what does this feature do to them? Not just the happy path user. The driver who just had an accident. The worker who depends on this income to feed their family. The person with no fallback.
A concrete example of what this could look like: a gig platform that shows drivers a real-time view of their effective hourly rate, after fuel and time, not just per trip. That one feature alone changes the power dynamic. It gives workers information they need to make real choices. It costs almost nothing to build. But it requires someone in the room to ask for it.
That someone can be you.
If you’re a designer, a PM, a founder, or an engineer, you are not just building features. You are writing the rules of someone else’s livelihood. That’s worth sitting with. Here’s a concrete checklist you can start using from today.
Who is the most vulnerable person in this system? Not the happy path user. The driver who had an accident. The worker with no fallback. Design for them too.
Who absorbs the cost of this efficiency? Every feature that saves the company time or money shifts cost somewhere.
Does this feature give workers more information or less? Transparency is a design choice. Hiding effective hourly rates, ratings logic, or deactivation reasons is also a design choice.
Can the people most affected by this decision push back? If a driver, contractor, or user has no way to appeal, contest, or opt out, that’s a power imbalance you built.
What behaviour does this incentivise at the edges? Algorithms optimise for what you measure. Ask what someone desperate, exhausted, or cornered would do inside your system.
Would I be comfortable if the people this affects could see exactly how it works? If the answer is no, that discomfort is information. Sit with it before you ship.
Are we baking in the ethics now, or leaving it for the legal team later? By the time lawyers are involved, millions of people are already inside the system. Ask the hard questions at the design stage.
Is this reversible if we get it wrong? Design for flexibility. Avoid lock-in that makes course-correction expensive.
Whose voice is missing from this decision? If the people most affected by this feature are not in the room when it is designed, go find them.
So ask yourself: in the last product decision you made, who benefited and who absorbed the cost?
Design is a force for good. But only if we choose to use it that way.