How Kickstarter Designed for Creators’ COVID-19 Concerns

The global pandemic spooked creators. We demonstrated the value of our platform by improving our ability to connect backers to projects.

As lockdown began, Kickstarter kept a close eye on how the pandemic might affect our ecosystem. We were focused on two primary indicating metrics: the number of launched projects from creators, and the overall pledge volume from backers.

Initially, the effects were minor. We saw a slight dip in launched projects, while pledge volume stayed somewhat steady. As April and May rolled around, we began to hear concern from creators both online and writing in to us about whether funding was drying up, or if they should delay their projects. Though backer support remained consistent, by May we had seen a 40 percent drop in projects funding on the platform — and no signal for the decline to halt.

May through July 2020 proved especially tough for the company, as we underwent a difficult round of layoffs and scaled our product organization down from six feature teams to two. With significantly less resourcing, our strategy for resiliency and stability had to be significantly more focused than in previous years. As a staff product designer with a strategic focus, my role was to inform the strategic prioritization process with our leadership team, as well as partner with Ellie Clinton, our team’s product manager, to identify how our team would contribute toward the company-wide objective of increasing support for projects seeking on our platform — without compromising on our charter principles.

Evaluating different pathways to resilience

With a smaller, more focused product organization, we began to chart our course. Before landing on a path forward, we explored a broad spectrum of strategies that could bring financial stability to the platform. We discussed everything from experimenting with our business model, to improving our creator tooling.

Ultimately, we felt that the best thing we could do for our creators was show them that creative projects were still finding funding on our platform. Terry Van Duyn, a group product manager here at Kickstarter, recently shared a quote from Marty Cagan that seems particularly relevant:

Early on at eBay we found we needed a product principle that spoke to the relationship between buyers and sellers. Most of the revenue came from sellers, so we had a strong incentive to find ways to please sellers, but we soon realized that the real reason sellers loved us was because we provided them with buyers. This realization led to a critical principle that stated, "In cases where the needs of the buyers and the sellers conflict, we will prioritize the needs of the buyer, because that's actually the most important thing we can do for sellers."

-Marty Cagan, Inspired

Leveraging our backers to foster a healthy ecosystem for funding creative work

In order to understand how we might drive more backings on our platform, our first step was to audit our existing backer journey.

Creating a backer journey map

Ellie and I dug into our data and analytics to uncover a few key performance metrics that informed our strategy:

  • 30 percent of searches on our site resulted in no results.
  • Our discovery page had a click-through rate of 10 percent.
  • Our project page had a one to two percent conversion rate.
  • Recommendations were not reaching backers.

The above stats show optimization opportunities across our home, search, discovery, and project pages, as well as our email communication. While we had some hypotheses as to how to improve performances in these areas, we conducted some user research to gain better insight into backer motivations.

Identifying hurdles in the backer decision making process

We conducted 10 total interviews with two cohorts of backers: experienced backers who had backed 10 or more projects, and new backers who had just backed their first project. We ask them to navigate to our home page, find a project they like, and walk us through their process of deciding whether to back a project while navigating a project page.

This was the first study of its kind in recent history, and it validated many hurdles we had long suspected: The backing process was fragmented, non-linear, and frequently veered off-platform. We were able to structure the key phases of the backer decision-making process as follows:

  • Learn the basics of crowdfunding
  • Get a quick overview of the project
  • Evaluate the risk associated with backing the project
  • Select a reward that meets their criteria
  • Understand next steps after backing

This new clarity around the backer decision-making process both enabled us to ship impactful improvements to our backer journey, as well as opened up new questions for us to explore.

Improving our ability to connect backers to projects they love

Equipped with a fresh perspective on ways to remove friction from the backing experience, we began ideating on different tactics and points along on the journey they might be implemented.

🔍 Search and discovery

  • Update our ElasticSearch to reduce the number of searches with no results
  • Improve the sorting algorithms that power search and discovery
  • Improve the usability of our search UI
  • Update our project card components to show more relevant project info

✏️ Project page

  • Give a stronger project overview by surfacing different info in the header
  • Highlight the creator’s relevant past experience running projects
  • Allow creators to add images corresponding to each reward tier

🎨 Recommendations

  • Email backers when a creator they’ve backed launches a new project
  • Include more recommendations in project updates emails
  • Introduce recommendations when a search has no results
  • Add recommendations at the bottom of project page
  • Increasing the number of recommendations in our carousel
  • Add recommendations to our backed projects page

Focusing our tactics

Having generated a robust list of potential tactics, we used RICE prioritization, staffing constraints, and a single cycle timeline to prioritize projects. To ensure a healthy ROI based on our team staffing and timeline, we set a key metric for our features to increase pledge volume by $1 million per month.

High priority tactics

We identified spreading the reach of our recommendations as a low lift, high impact sweet spot.

  • Add recommendations at the bottom of project pages
  • Email backers when a creator they’ve backed launches a new project
  • Include more recommendations in project updates emails
  • Introduce recommendations when a search has no results

Follow up tactics

As fast follow ups to our high priority projects, we prioritized improvements to our design system at the component level that would improve engagement rates by surfacing more relevant information to backers.

  • Give a stronger overview of a project in the project page header
  • Strengthen the algorithms that power search and discovery
  • Update our project card components to show more relevant info

📈 Impact summary

  • We set our key result as gaining additional $1 million per month, and surpassed double that number. We’ve seen between $2.2 million and $2.9 million per month since July.
  • Revenue from these features represented nearly two percent of our total annual pledge volume in 2020, and is on pace to represent three to four percent in 2021.
  • 46 percent of all backings in November 2020 were driven by Kickstarter-referred sources, the highest in the history of the company.

🔑 Tactical successes

  • The Creator You’ve Backed email generates ~$1 million in pledges per month. We collaborated with Meg Heim, our senior director of Integrity, to ensure these emails are only sent for creators who have high sentiment scores from their backers. The strong performance of this email validates the direct correlation between backer retention & project sentiment.
  • Recommendations in project updates generate ~$1 million in pledges per month. While our recommendation engine is young and has opportunity to grow, we were able to significantly boost its performance by making recommendations more discoverable.
  • Recommendations at the bottom of project pages generate ~$500,000 in pledges per month. Before scaling to a full launch, we released this as an Optimizely experiment to 10 percent of users to ensure that there was no negative impact to conversion for creators.
  • Swapping our default Search and Discovery algorithms from “Magic” to popular increased click-through rate by three to five percent.

🧠 Learnings and looking ahead

  • Adding recommendations to searches with no results adds negligible revenue. This tells us that we need to invest further in ElasticSearch and sorting algorithms, as well as potentially improve the usability of our Search UI.
  • Upon doing engineering spikes into our project page and project cards, we uncovered significant amounts of technical debt that prevented us from being able to execute on those tactics within our timeline. These findings informed our vision through 2024 for how we will work to pay off tech debt.
  • We validated that factors that motivate backers to support projects are far more complex than what our recommendation engine currently takes into account, which is primarily the project category. Backers are also interested in how experienced the creator is, if the project supports a social cause, the price range of the reward tiers, how many backers the project has, when the rewards are expected to be delivered, and many more. This will fundamentally inform how we collect backer preferences moving forward.

Staff Product Designer @ Kickstarter. www.arjunmahesh.com

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store