Many products function as systems where users aren’t just consumers but active participants: they post, review, report, vote, contribute, or collaborate. The whole thing only works because people choose to lean in. And whenever a team wants to increase that participation, the first suggestion is often: “Maybe we should add an incentive.”
Incentives do work. They just don’t always work the way you expect.
During my Master’s studies in HCI, I ran experiments on a community platform I built to understand what actually motivates people to contribute. We tested various prompts and reward structures—from impact-framed messages to straight-up cash incentives—and watched how different kinds of people reacted.
The first thing that stood out was that users weren’t interchangeable: the same nudge could energize one person, annoy another, and barely register for a third.
The sharpest pattern was about incentives. Financial or reward-based prompts created a fast spike in activity but also the fastest drop-off. They reliably got people “in,” but they also made people feel like they’d done their part once they’d collected the reward. The behaviour shifted into something transactional, and the system started to feel more like gig work than participation. In the live deployment, you could see it clearly on the timeline: the cash group surged in the first couple of days and then fell off a cliff faster than any other condition.
By contrast, prompts that focused on impact, community, or competence activated more slowly but led to steadier engagement. People contributed because they felt their input mattered, not because they were chasing something. The quality improved, discussions were better, and people were more likely to return without additional nudging.
None of this was theoretically new—Self-Determination Theory had predicted these patterns for decades—but the interesting part was seeing how cleanly they mapped onto a real product. Different motivations, different users, different outcomes.
The practical insight here is that your user base isn’t one blob of “motivated people.” It’s a portfolio of motivational profiles. And the moment you treat it like a single segment, you distort the system.
In the live deployment, the interaction effects were surprisingly asymmetric:
- Control-oriented users, when shown a matching incentive (e.g., a cash raffle), showed about 40% more interactions than their baseline.
- Autonomy-oriented users, when shown that same cash incentive, dropped to roughly 45% fewer interactions than their baseline.
- Meanwhile, control-oriented users shown only autonomy-style “do it for the community” messaging also reduced engagement.
In other words: incentives aren’t neutral. They’re highly contingent on who you’re showing them to. A well-matched nudge accelerates the system; a mismatched one poisons it. This is also why so much gamification fizzles out—reward-heavy mechanics work at first but age badly, especially for the autonomy-driven segment.
A healthier approach is to think in terms of a motivation portfolio. Mix and match the levers: impact cues (“your input improves the experience for everyone”), competence cues (“you helped solve X”), community cues (“others benefit from what you share”), and yes, the occasional strategic incentive. Rewards are accelerants, not foundations.
Product & Design Principles
-
Segment by Motivation, Not Demographics.
The biggest difference between users usually isn’t age or location—it’s why they participate in the first place. -
Deploy Incentives Surgically.
Rewards work well for your control-oriented users and for kickstarting new loops, but they lose power fast and shouldn’t be the backbone of a participatory system. -
Avoid Reward Contamination.
Don’t rely on financial incentives on autonomy-driven users. For them, rewards can create a measurable collapse in engagement. -
Design for Long-Term Autonomy.
The sustainable core of any community rests on autonomy-supportive structures: showing impact, showing competence, showing connectedness. -
Treat Contribution Like an Ecosystem.
Participation isn’t a funnel—it’s a living system. Nudge it carefully, or you distort it.
Ultimately, designing for participation means shaping long-term dynamics, not just juicing next week’s numbers. The systems that thrive are the ones that understand why people show up in the first place—and avoid killing the vibe by paying for something that didn’t need payment to begin with.
References
- Personalized Motivation-Supportive Messages for Increasing Participation in Crowd-Civic Systems. P. Grau, B. Naderi, J. Kim. In Proceedings of the ACM Conference on Computer-Supported Cooperative Work (CSCW), 2018. Download
- Why We Do What We Do: The Dynamics of Personal Autonomy. E. Deci, R. Ryan. 1996.