In December, Prisma - an ORM for Node.js and TypeScript - had 3,300+ open issues. And Prisma acknowledged things weren't great with how they handled these issues:
"we’ve faced challenges in governance, issue management, and communication. Priorities haven’t always been clear, deadlines haven’t been consistently met, and over time, we’ve accumulated 3.2k open issues and a backlog of preview features stretching back years."
Prisma is a hugely successful open source project - it powers 500,000+ repos and has 400,000+ monthly npm downloads.
And the reality is that the Prisma team was working really hard: they had closed more than 7,500 issues in recent years.
But it wasn't enough. With their popularity, came a huge weight of community requests that were impossible to keep up with without making changes.
In this newsletter, I summarize how Prisma completely transformed their community sentiment with transparent issue prioritization.
Enter the ORM manifesto
In December, Prisma released an article titled: "Prisma ORM Manifesto: Clarity and Collaboration".
The document lays out in detail:
1) Which databases they will prioritize
They explicitly defined which databases they will prioritize and fully support moving forward: PostgreSQL, MySQL, SQLite, MongoDB, and MariaDB.
2) How they will manage issues
Organization:
- Aggressively close outdated/addressed/unaligned issues.
- Label issues and set timelines
- Recommend that any closed issues that are relevant a team/org move to the enterprise support plans
Prioritization:
- Th most upvoted and commented-on issues are prioritized first
- First class database related issues take priority
AI automation
They will use Dosu to "craft thoughtful responses, maintain effective engagement, and resolve more issues on GitHub."
3) How they move from preview to GA
Moving forward, any feature released to Preview in a given quarter will reach General Availability (GA) the following quarter.
4) TypeScript + more focus on extensibility
Prisma has migrated its core functionality from Rust to TypeScript. This transition enables Prisma's core community to contribute and extend Prisma.
Prisma also shares their guiding principles:
I interviewed Søren, the CEO of Prisma this week and he reiterated they prioritize features based on:
- Their company's strategic initiatives
- Prisma community's upvotes and comments
So, if someone wants a feature or bug worked on, Søren himself welcomes asking other Prisma users to brigade the issue.
Get your friends to come and upvote. And that works because then we work on the thing.
He also shares that this has completely transformed Prisma:
"So silly, so simple but we've been doing that for year and it has completely changed the tone and the general feedback we got on twitter and social media. There is this perception that Prisma is now much more responsive, receptive to community requests and moving faster"
Despite keeping the size of the team the same, by changing their approach to prioirtization, the features they work on now impact many more people and many more users feel "oh, they implemented a feature I wanted", because they are working on the highest impact features.
Today, only a month or so later, Prisma now has only 2,700 open issues - a reduction of 500 issues.
Don't build more elevators
There is a famous story that a hotel was receiving complaints about slow elevators. Correcting this would be extremely expensive.
But then someone smart had an idea to put mirrors in elevator. This gave the elevator passengers something to do and the complaints stopped.
The core issue was not slow elevators, it was bored elevator passengers.
It's clear that Prisma followed the same problem solving approach: when faced with more GitHub issues than they could solve, it would have been easy for Prisma to say "we need more VC money and more engineers and then we can close more GitHub issues".
But they realized the core problem was not "there are too many issues", it was "the community feels they aren't being heard".
So they narrowed the scope of their core community, they communicated clearly Prisma's priorities and they gave agency to the community (upvotes and extensibility).
If you find this interesting, check out the full interview with Søren. We also talk about why they released their database product and why they don't use AWS.