Can we nudge people to earn extra income?

Interventions
In-app framing
Experiment Type
Field Experiment
Key Behaviors
Find additional employment, Increase income or sales
Outcomes
Increase earnings
Focus Areas
Marketing & messaging
Behavioral Concepts
Framing Scarcity mindset
Partner
Steady
Partner Type
Fintech/tech
Collaborator
Irrational Labs

What Happened

It didn't work. There was no significant difference in Income Booster completion among the eight framing conditions.

Lessons Learned

Changing the framing of the Income Booster screen alone has not proven to be an effective method in increasing users' willingness to enroll in new jobs/earn extra income.

Background

For those currently working in the gig economy, nearly half (44%) depend on gig work as their primary source of income. Steady is an app that helps these workers maximize their income. They do this by matching people with opportunities to make extra money and providing an income tracker to monitor earnings.

One of the ways people can earn extra money is via bonuses (or Income Boosters) that Steady gives out when their users sign up for new jobs or services. For example, a Steady user could earn $20 when they sign up to be a delivery driver for Postmates. However, not all Steady users take advantage of the Income Boosters program. Steady partnered with Common Cents Lab to explore ways that we might encourage greater adoption of these bonus programs.

Key Insights

To better understand what might keep Steady users from pursuing these bonus programs, we conducted a detailed diagnosis of the Steady app flow. We identified several barriers that we believe influence people’s decisions to act on income boosters.

  • The Income Boosters lack a clear deadline. We know that people find procrastinating much easier when tasks are left open. Adding a deadline also creates a degree of scarcity, which focuses attention and increases the likelihood someone will act.

  • The Income Boosters feel like a generic part of Steady’s offerings. Services that are widely available and that don’t feel tailored are easily undervalued and ignored.

  • There were no clear signals for whether or not other Steady users use Income Boosters. People often look for cues about what others are doing to guide their own decision-making.

Experiment

In order to increase the number of users who pursue income boosting opportunities, we identified and leveraged key behavioral principles that could drive decision making. CCL and Steady designed a 4x2 experiment to test which approach would be the most effective in maximizing user engagement with Income Boosters. Approximately 36,000 unique Steady users received one of the eight randomized experimental conditions the first time they reopened the app after their initial registration with Steady.

Results

We measured the difference between the initial acceptance rate (clicking: “Boost my Income!”) and the “click out” rate. The “click out” rate was the next step for users to claim their money.

With regards to the initial acceptance rate, there was no statistically significant difference in the two overarching frameworks, “Transparency” and “Bonus Cash.” Both conditions averaged a 42% initial acceptance.

However, across both “Bonus Cash” and “Transparency,” the deadline approach resulted in the highest CTA acceptance rates (45.3%). The exclusivity framework resulted in the lowest (35.3%).

We found no statistically significant difference in click-outs across all conditions. Despite variance in CTA acceptance rates, none of the conditions resulted in an increased Income Booster completion. Why no difference? There are a few possible explanations:

  • Almost one-third of Steady users took advantage of the income-boosting opportunity. By most mobile app standards, this is considered very high engagement. It’s possible there just wasn’t an opportunity to move this number even higher, even with the compelling deadline.

  • More likely, the ask (“sign up for a new service!”) could be perceived as very difficult, requiring a lot of work. The deadline nudge simply may not have been strong enough to overcome the perception of friction that was implicit in this ask.