Getting Started

Gitcoin is a data rich platform that is moving steadily towards leveraging its treasure trove of information in novel and interesting ways. This analysis of the lifetime Gitcoin grants data seeks to be both comprehensive and insightful with just enough quirkiness thrown in to highlight just how interesting the analysis of this data can be.

Grant Round Timing

First, we can visualize the active grant round periods over time. One would expect regular intervals of time between each round, but that historically hasn’t been the case.

Round Start Date End Date
1 2019-02-01 2019-02-15
2 2019-03-26 2019-04-19
3 2019-09-15 2019-10-04
4 2020-01-06 2020-01-21
5 2020-03-23 2020-04-05
6 2020-06-16 2020-07-03
7 2020-09-14 2020-10-02
8 2020-12-01 2020-12-18
9 2021-03-10 2021-03-25
10 2021-06-16 2021-07-02
11 2021-09-08 2021-09-24
12 2021-12-01 2021-12-16

We can make this more clear by stacking each year vertically, which indicates that the grant rounds both occurred at different times during the year, and lasted for different amounts of time with 2020 being one of the busiest years.

Text Analysis on Grant Titles

With a little bit of text analysis we can better understand the types of grants that are hosted on the site by categorizing them according to their titles. Here we’ve performed a Latent Dirichlet Allocation algorithm on the grant titles, after performing some basic text cleaning operations. We selected six themes for the analysis. Below we can see the top terms that fell into each of the six categories.

Roughly speaking, the categories obtained for the grants are as follows:

  1. DeFi - Grants associated with Decentralized Finance
  2. Blockchain/Crypto - General blockchain development projects
  3. Panvala League - Grants associated with the Panvala shared endowment
  4. DAO and Community - Grants seeking to build a community or DAO
  5. Insights and Media Awareness - Grants involving dashboards and social media awareness for the project
  6. NFTs - Grants related to NFTs

A fun exercise is to look at the amount of money raised by projects with specific words in their descriptions. Not surprisingly the word “blockchains” was used in quite often and the total amount raised by grants with that in the title amounted to $6,984,782 across all rounds. Other popular words with donors were “policy”, “makers”, “privacy”, and “accounting”.

Top 30 words by total amount raised

word total
blockchains $6,984,782
policy $5,889,070
makers $5,889,034
privacy $5,007,185
accounting $2,911,117
protects $2,910,448
rotki $2,910,448
prysm $2,730,737
prysmatic $2,730,737
poap $2,515,437
attendance $2,430,478
lighthouse $2,299,218
dark $2,251,306
simple $2,107,494
ethers.js $2,105,853
tiny $2,105,853
hardhat $1,748,456
information $1,531,727
ethhub $1,529,847
commons $1,502,473
brightid $1,479,451
uniqueness $1,479,451
nomic $1,367,364
forest $1,365,661
team $1,364,401
transactions $1,321,800
nimbus $1,300,836
zeropool $1,299,591
payments $1,286,155
preserving $1,266,625

An slightly deeper analysis normalizing for the word frequency could further unearth other specific words that resonate with Gitcoin donors enough for them to donate heavily to grants with those keywords. An even more interesting follow up would be to see how the previously derived topic-model based categories have changed over time across rounds. This could answer questions like: Have there been more or fewer social awareness grants? or When did NFT related grants become popular?

Distributions and Spreads

We now want to look at how the monetary contributions are spread out by contributions, the number of unique contributors and matching funds. The most unsurprising results we see is that the most common contribution is indeed $1 and was, by far, the amount that appeared most frequently in the data. On average donors gave somewhere between $100 and $500 as indicated by the vertical line. At the skew we see about some contributions totaling more than $10,000. Across the rounds you can see not only the total volume increasing, but the average donation moving upward overtime.