Cutting Six Scrolls to One Glance, Where a Wrong Schedule Means Nobody Gets Paged
Area : B2B Saas , Incident Management, BLR IND
When : 2022 - 2025
Role : Product Designer
TL DR
The incident list was a doorway, not a workspace. I turned it into the place engineers could actually act
I redesigned the first touchpoint of incident response, the list view, so on-call engineers could acknowledge, reassign, and reprioritize in one click without opening a single incident. The goal was simple: shrink the time between seeing an incident and acting on it, because at scale those seconds are downtime you pay for.
Problem
It started with our CEO being OnCall He told me acting on incidents from the list was painful sometimes because of the constant back and fourth The list wasn’t a workspace. It was a doorway you had to walk through every single time.
Acknowledge
Reassign
Change priority
Snooze
Resolve
Acknowledge
Reassign
Change priority
Snooze
Resolve
How can we make incidents more visible to engineers during the most critical moments, so they can acknowledge and respond faster
This became the brief
20
Conducted User Interviews with 20 OnCall Engineers
140
Collected 140 Session recordings to map user journey
5
Created a group of 5 Internal testers from different OnCall teams
So I observed for a few days internally while the users were managing incidents to learn about their core workflow and found out
The list showed almost nothing, so engineers opened every incident just to triage it
Click are was the incident number making it super difficult to click on the incident
Every action forced a context switch. Open, act, back out, repeat.
The things people reached for most took the most clicks to reach.
Core incident actions (Acknowledge, Resolve, Snooze) made instantly accessible from the incident list
Inline tagging introduced for faster incident organization
Incident title visibility increased to support up to 120 characters without truncation
Filters and relationship indicators introduced for collated incidents
Confirmation workflows added for critical actions like Acknowledge, Resolve, and Snooze
Bulk Acknowledge and Resolve actions introduced for faster incident management
Impact After the redesign
120%
Higher Incident Visibility on Incident Title upto 140 chars
12 Clicks
Removed from common actions related to the Incident
8 Actions
Made available in the Incident List itself
12 Sec
Average Reduction in MTTA
In Numbers
A 40 to 50 second MTTA is generally considered good for highly automated SaaS teams, although elite SRE organizations often target under 30 seconds for critical production incidents.
Many organizations still operate in the 1 to 5 minute range depending on their maturity and on-call processes. Razorpay processes around $180B in annual payment volume and generated ₹3,783 Cr ($450M+) in revenue in FY25.
If a company that size sees ~50 production incidents a month, shaving 10 seconds off acknowledgement gives back over 8 minutes of engineering response time every month.
At industry estimates of around $2,000 per minute of downtime, that's roughly $16,000 (₹13-14 lakh) in operational impact. More importantly, those seconds can prevent customer disruption during critical payment flows.
