Lean ChatOps for L3 support

Introduction

As payment engineers at SwissBorg, one of our core responsibilities is assisting with various operational requests coming from payment agents. These requests can range from simple inquiries to complex technical issues, and it’s crucial that we track their frequency and distribution among our team members. Our team primarily communicates through Slack, and we needed a way to manage these requests with a lean and lightweight process that wouldn’t bog down our operations. Creating Jira tickets for every request simply wasn’t feasible, so we had to come up with a more efficient solution.

Solution: A Simple Workflow Built on Slack Workflow Builder

To address our needs, we designed a straightforward workflow using Slack Workflow Builder that allows us to manage operational requests efficiently and effectively. Here’s how it works:

1. Reaction-based triggers: Our workflow is activated by adding a designated emoji to any message in our support channel. This makes it easy for our team members to raise an issue quickly without creating a separate ticket.
2. Issue prioritization: Once the workflow is triggered, it automatically prompts the message author to assign a priority level to the issue. This helps us focus our efforts on the most urgent tasks first.
3. Engineering channel notification: After the priority level is set, the issue is posted in a separate channel dedicated to engineers. This keeps our support channel clutter-free while ensuring that the right people are notified about the issue. It also gives us a place where we can discuss technical matters without bothering agents.
4. Claiming requests: In the engineering channel, team members can easily pick up requests by clicking a button inside the message. This not only allows for the seamless allocation of tasks but also prevents duplicate work on the same issue.
5. Request author notification: As soon as a team member picks up a request, the original author is notified. This provides transparency and reassurance that their issue is being addressed promptly.
6. Google Spreadsheet for tracking: To maintain a record of all the requests and their details, we save this information in a Google Spreadsheet through Slack Workflow integration. This allows us to monitor trends, analyze the load distribution among team members, and make data-driven decisions to optimize our support process.

How does it look?

Here you can see an example of interaction with the flow:

And here is the workflow definition:

And this is the view from engineers' perspective:

Simple, isn't it?

Wording matters

An essential aspect of our solution is the way we ask about the urgency of the issue. We understand that the language used in this step can significantly impact the effectiveness of the entire process. Therefore, we have devised a clear and specific method for prioritizing requests:

• Avoiding ambiguity: Instead of using labels or numbers that can be open to interpretation, we opt for a more descriptive approach when asking about the priority of a task.
• Three clear options: To keep the process simple and prevent decision paralysis, we provide three priority choices:
  1. Best if done immediately
  2. Best if done today
  3. Best if done within a few days
• Fostering psychological safety: We phrase the priority options with a “best effort” mindset. This approach ensures that the agent feels comfortable selecting their desired priority level without worrying that they might distract the team. We don’t enforce strict SLAs; if no one is available to address the issue within the desired time frame, it won’t cause unnecessary stress or repercussions.

When Simplicity Isn’t Enough: Considering More Sophisticated Solutions

While our lean chat ops solution has been successful in streamlining our L3 support process, it may not be able to meet more complex requirements. The workflow we’ve designed focuses on simplicity, and this might not be sufficient for all situations or organizations.

In cases where our current process falls short, we would consider embracing more sophisticated solutions, such as Atlassian Halp or Jira Service Management. These platforms offer advanced features and customization options that can cater to a wider range of support scenarios and organizational needs while still allowing for Slack-first interactions.

However, our goal remains to maintain the simplicity and efficiency of our current process for as long as possible. We hope that the need for more complex solutions won’t arise, allowing us to continue leveraging the advantages of our streamlined Slack-based workflow.

Conclusion

By implementing a lean chat ops solution using Slack Workflow Builder, we have significantly streamlined our L3 support operations at SwissBorg. The simple workflow has not only improved our team’s efficiency but also allowed us to track and analyze support requests with ease. This approach has proven to be a game-changer for our payments team, enabling us to better serve our payment agents and continuously enhance our support processes.