Wolt is a technology unicorn with 3500+ employees and 200+ engineers building "the everything app" for goods and services, operating in 23 countries. For the past five years, Wolt has been scaling tremendously, creating a demanding environment for a growing engineering organization. Wolt is a perfect example of how Swarmia creates value for modern high-scaling engineering organizations with greater transparency, higher productivity, and increased coordination.
One of Swarmia's leading principles is that every team is different. This applies well to Wolt's engineering organization, where a team working on a core product feature will have very different problems from a team working on a greenfield project. Swarmia has brought Wolt a scalable way to remove blockers in each team, fueling software excellence in the whole organization.
We discussed with a team working on one of Wolt's core features to highlight their problems and how they use Swarmia to solve them. They are not following any strict process, but work resembles mostly Kanban. The team is scoping epics for project-like things in Jira. Together, they had concluded that they have three main challenges:
- Pull requests are sometimes falling through the cracks
- Too much multitasking
- How to ensure allocating enough development time into technical housekeeping
With Swarmia, the team has solved all of these problems.
1. Improved visibility into pull requests in progress and introduced cycle time target to reduce cycle time 20% from 5 to 4 days by reducing outliers
To ensure that no single pull request is forgotten, the team follows Swarmia pull request view to identify and move forward any stale and old pull requests. They have also adopted a Working Agreement to close pull requests in less than 7 days. With the Working Agreement in place, they get automatically notified in Slack about any exceptions to their target. Individual developers use Swarmia Slack notifications for pull requests to get pinged when a review is assigned or when their pull request has been approved or failed CI checks.
2. Reduced multitasking with WIP limit resulting in 26% increase in throughput
To increase transparency into multitasking, the team uses Swarmia's Work Log view to visualize where they are spending their time at any given week. To prevent drifting towards multitasking, they adopted a Working Agreement to limit work in progress. With the Working Agreement, they are always notified when they break their work in progress limit and can focus on closing work before starting any new work.
3. Surfaced insights into resource allocation with Investment Distribution view
The team had created an epic in Jira to follow any work categorized as technical housekeeping. However, it was still painful to follow if they hit their target level on resources spent into this epic. The Investment Distribution view allows the team to track how much time they spend in housekeeping, based on visualizing Jira data. This means the team doesn't have to second-guess how they're doing against their targets – but instead, they can get an answer right away and communicate this also elsewhere in the organization.