Incentivizing user generated content: gift cards vs. donations

CASE STUDY

Highlights

• G2's founders established an initiative called G2 Gives to provide donations for non-profits who align with G2's values.
• G2 regularly incentivized reviewers with gift cards, but product leaders wanted to see if we could shift more money toward donations with a "swap" campaign.
• Using A|B tests we determined the best performing version.
• We also gained insights about a correlation between user motivations, quality content, and likelihood to switch to a donation.
  • People seeking methods of income by writing content look for paths of least resistance (and ways to "game the system") and choose to maximize gift cards over donations.
  • Their rewards are not quality content but monetary payout.
  • People motivated by sharing with the community the story of their experience in hopes of helping others navigate the same situations are also more likely to contribute quality and to respond to invitations to be generous in other ways.

Contents

1. Setting the stage
2. Identifying assumptions
3. Design explorations
4. Prioritizing solutions
5. Impact
6. Insights about motivation and quality

Setting the stage

At surface level this was a straightforward initiative—new feature implementation plus iterative optimization. The desired outcome was generally easy to understand—maximize donations by converting reviewer incentives into donations. At a deeper level, however, there was a lot of nuance, complexity, and psychology to navigate.

We needed a nuanced UX content strategy we could prototype and optimize.

• Would emphasizing that this was their earned reward encourage them to have agency to use it as a donation, or would that create a sense of personal loss?
• If the donation was of a higher amount would that encourage them to lean toward the donation to generate more value in the world—"Convert a $10 reward to a $15 donation"?
• What images would encourage optimizing for donations? What content, if any, would distract from the action we hoped to encourage?

Identifying assumptions

Product leaders who believed in the power of G2 Gives and the valuable work of the charities we partnered with believed that many users would favor "doing good" over pocketing the small rewards G2 gave for completing review content.

They also believed that the right inflection point for the conversion ask was immediately following the review submission when we normally provided a link to the user's gift card.

In the first iteration of the "donation swap" feature, we believed demonstrating impact through New Story would create a sense of well-being in users who chose to partner with G2 and New Story to build homes for families in Haiti or Mexico.

The initial assumption was that providing the reward funds and then inviting a donation as an either/or choice would result in the most donations.

Phrasing these assumptions as questions that could be tested, we set out to discover answers that would influence our designs:

• Are users willing to convert reward funds into donations? What factors encourage this decision and what factors discourage it?
• At what point in the user generated content flow are users most likely to swap their reward for a donation?
• What impact does demonstrating impact have on the likelihood of a donation?
• What aspects of messaging most impact donations over rewards?

Design explorations

Leveraging our design system components for the most efficient first iterations, I focused on text and call to action variations, lining up the either/or choice product believed would be most effective. I included the increased donation amount to gain insight on that dynamic from user feedback.

Messaging was conversational and attempted to avoid "shaming" a person into a donation.

We represented the rewards with our usual Amazon gift card image and the donation with an image from New Story that we hoped would express joy and celebration at generosity.

Testing this prototype with users we learned a lot to correct our assumptions.

With this prototype, users felt we had "handed" them a gift and then were asking to take it back from them. This confused some of them. They ended up having a lot of questions at a point in the user flow where they expected to receive a simple "thank you" gift. The ask did not align with their expectations.

The donation amount increase also confused them and some wondered if that was a typo; we didn't explain the increase well in the content.

Phrases such as, "...if you like..." undercut the potential clarity of the decision presented to the user.

Although leaders at G2 understood the integration of G2 Gives with G2, the existence of G2 Gives was news to many users and they had questions about the program before they felt comfortable supporting it. Cognitive load was high.

This data led to several other rapid prototypes.

Prioritizing solutions

The iteration we finally implemented took an "announcement" approach for the invitation to donate. "Good news! You can transform your reward into a donation!" replaced the idea that we wanted to "take" their reward for another purpose.

This approach also afforded us the opportunity to introduce the swap decision at several different points in the flow, including through a congratulations email once a review was approved.

We kept the reward and donation amounts the same to minimize confusion and added a little more information about New Story and the particular initiative they would be contributing to.

The call to action button to donate was primary, and we added the idea that other users were donating as a way to encourage them to expand their thinking beyond the immediate gratification of the reward and join their peers.

The question, "How many can we build?" was a gentle challenge to encourage a mindset of incremental contribution to a community adding up to bigger impact.

Lastly, the "No thanks" button was our tertiary button style so that it was present but clearly not the preferred action for this screen.

We tried a number of different images and this one resulted in the most positive feelings from user tests.

Impact

The swap campaign feature was successful in achieving the outcomes desired. In fact, we increased our goal to encourage more donations through the end of the year.

This campaign helped G2 provide a total of $47,770 to New Story to build almost eight homes ($6,000 each).

Insights about motivation and quality

These optimizations kept me looking at the data and asking questions about possible patterns in it.

One insight that emerged from the data was that there was a correlation between those who swapped a reward for a donation and review quality scores. It was not a perfect correlation but still a noticeable pattern.

Through user interviews I was able to understand that not all users were motivated by the gift card rewards. There are some who would contribute review content without any monetary incentives because their motivation was to share their experiences and stories with the community. Those users also tended to have higher quality scores because they shared more information in their reviews. They also tended to donate when invited to do so. Their reward was sharing with the community and helping others—whether through their reviews or swapping their gift card for a donation.

In contrast, users who were motivated by the promise of a gift card tended to look to maximize the number of gift cards they could get and to do the least amount of work to maximize their reward. These reviews were of lower quality and review submission volume was higher.

1. People seeking methods of income by writing content look for paths of least resistance (and ways to "game the system") and choose to maximize gift cards over donations.
2. Their rewards are not quality content but monetary payout.
3. People motivated by sharing with the community the story of their experience in hopes of helping others navigate the same situations are also more likely to contribute quality and to respond to invitations to be generous in other ways.

These insights triggered future explorations of how we could adjust our reward systems to shift away from heavy use of gift cards to other forms of reward that would improve quality user generated content—discovering ways to build a sense of community through features like peer-to-peer interaction, public discussions and expert user profiles.