Characteristics of On-Time and Late Reward Delivery Projects

Thanh Tran, Kyumin Lee

TL;DR

While most crowdfunding research focuses on how projects get funded, this paper investigates what happens after the campaign: reward delivery. By analyzing over 2,000 Kickstarter projects, the authors identify key project, creator, and linguistic characteristics that distinguish on-time deliveries from late ones. Ultimately, less complicated projects with highly responsive creators are much more likely to ship rewards on time.

What Problem Does This Paper Address?

Reward-based crowdfunding platforms like Kickstarter have exploded in popularity, but they suffer from a major fulfillment issue: roughly 35% of backers do not receive their rewards on time. While existing literature heavily focuses on the fundraising phase (e.g., predicting campaign success, analyzing backer dynamics), very little attention has been paid to the reward delivery phase.

Late deliveries damage backer trust and reduce the likelihood of future investments on the platform. This paper aims to bridge that gap by identifying the distinguishing properties of on-time versus late reward delivery projects. By understanding these characteristics, platforms can better assess risk, and creators can improve their project planning and backer communication.

Methods

The authors took a data-driven, statistical approach to compare on-time and late projects. They extracted four categories of features:

  1. Project characteristics (e.g., funding goal, number of rewards, description readability).
  2. Creator characteristics (e.g., past backed projects, bio readability).
  3. Activeness (e.g., update frequency, comment response time).
  4. Linguistic patterns (analyzing the text of updates and comments).

For the linguistic analysis, they used the Linguistic Inquiry and Word Count (LIWC) dictionary to map text from creator updates and backer comments into 64 psychological categories. They then applied one-tailed two-sample t-tests (using Bonferroni corrections to account for the multiple LIWC categories) to find statistically significant differences between the on-time and late groups.

Datasets

  • Custom Kickstarter Dataset: The authors sampled 2,949 successfully funded Kickstarter projects launched between 2009 and September 2014.
  • Ground Truth Labeling: Because Kickstarter does not strictly track fulfillment dates, three independent labelers manually reviewed project updates and comments to establish ground truth.
  • Projects were labeled "on-time" (1,003 projects) if all rewards shipped by the Longest Estimated Delivery Date (LEDD) with no complaints, and "late" (1,195 projects) if creators announced delays beyond the LEDD. 751 projects were excluded due to unverifiable shipping statuses.

Key Insights and Ideas

  • Complexity is the enemy of punctuality: Projects with lower funding goals, less pledged money, and fewer reward tiers are significantly more likely to deliver on time. Massive funding success can actually be a double-edged sword if a creator is overwhelmed by the logistics of overfunding.
  • Category dictates delivery risk: Dance and theater projects (which are often live performances or classes) usually deliver on time. Conversely, technology, games, comics, and film projects (which require complex manufacturing and physical shipping) are highly prone to delays.
  • Communication cadence matters: Creators who post updates at shorter intervals and respond to backer comments quickly are strongly correlated with on-time deliveries. Silence or slow response times often precede a delay.
  • Linguistic "tells" in updates: The language used in project updates is a strong indicator of fulfillment status. Creators of late projects use significantly more "cognitive process" words (e.g., cause, know, ought), likely because they are constantly explaining why delays are happening. On-time creators use more positive and affective words.

Key Results

  • Project Stats: On-time projects had significantly lower average funding goals ($7,670 vs. $13,584) and less total pledged money ($14,419 vs. $30,126) compared to late projects. They also featured fewer images, videos, and FAQs.
  • Readability: On-time projects had lower SMOG (Simple Measure of Gobbledygook) scores in their project descriptions (average 30.12 vs 38.59), indicating simpler, more readable text.
  • Activeness: Creators of on-time projects had a significantly shorter average time interval between consecutive updates (p < 0.001) and responded to backer comments faster (p < 0.05).
  • Backer Sentiment: Backers of late projects asked significantly more questions about reward progress (p < 0.001). Furthermore, backers of late projects used more words associated with negative emotion and anxiety, mirroring the delayed status of the project.

Limitations and Future Work

  • Limitations: The study relies on the manual labeling of updates and comments to determine delivery status. This restricts the dataset size (2,198 usable projects) and makes scaling the analysis to the entire platform difficult. Additionally, the study is purely observational; it identifies statistical correlations but does not build a predictive machine learning model to forecast delays at the time of launch.
  • Future Work: While not explicitly detailed in a dedicated section, the authors frame this as the "first step" toward trustworthy crowdfunding platforms. The natural next step is using these extracted features to train predictive models that can flag high-risk projects early, or to build recommendation systems that guide creators on better project design and communication strategies.

Summary updated

Abstract

The crowdfunding market size has increased exponentially, reaching tens of billions of dollars and showing the popularity of crowdfunding. However, according to Kickstarter, 35% backers did not receive rewards on time. To maintain the trust between creators and backers, and sustain the crowdfunding business growth, it is crucial to understand how on-time and late reward delivery projects are different. In this paper, we analyze characteristics of on-time and late reward delivery projects, especially, focusing on project descriptions, creator profiles, and activeness and linguistic patterns of creators and backers. Our analysis reveals that the less complicated a project is and more actively a creator responds to backers, the higher an on-time reward delivery probability is. It shows there are significant differences between on-time and late reward delivery projects.
BibTeX
@inproceedings{tran2017characteristics,
  author = {Thanh Tran and Kyumin Lee},
  title = {{Characteristics of On-Time and Late Reward Delivery Projects}},
  booktitle = {ICWSM},
  year = {2017},
  url = {https://openreview.net/forum?id=3rZ4xi7DiZ},
}

Loading related content...

Figure