SafetyIQ Dashboard evolution

Highlight

30% more active users, +15 min/user/week engagement

Problem

Insurance PMs were dropping off after the spending a few minutes on Dashboard. Users weren't complaining but metrics were flat and feature requests weren't moving the needle.

Solution

Interviews revealed nothing obvious. Observation did users couldn't read the graphs.

  • Added AI summaries per graph translating data into plain language

  • Added AI page-level summary giving users the full picture instantly

  • Rebuilt information architecture to match real user workflows


Date

Dec, 2025

Watch the summary video to skip reading.

Watch the summary video to skip reading.

What is SafetyIQ

What is SafetyIQ

SafetyIQ is a fleet intelligence dashboard that analyzes driver behavior and safety events to help fleets understand risk patterns, improve driver performance, and make safer operational decisions for Insurance companies.

SafetyIQ is a fleet intelligence dashboard that analyzes driver behavior and safety events to help fleets understand risk patterns, improve driver performance, and make safer operational decisions for Insurance companies.

30+ Insurance companies

500+ Fleets

1000s of Drivers

Millions of Data points

Couldn't answer one question
What do I do next?

Summary

Summary

We conducted 2 separate Maze usability tests, 6 targeted users participated.

  1. AI Insights

  2. Program Dashboard (PD)

We conducted 2 separate Maze usability tests, 6 targeted users participated.

  1. AI Insights

  2. Program Dashboard (PD)

Objective

Objective

Position SafetyIQ as:

  1. An indispensable part of daily workflow with the new Dashboard re-design

  2. A critical tool for fleet risk assessment and prioritization with AI

Position SafetyIQ as:

  1. An indispensable part of daily workflow with the new Dashboard re-design

  2. A critical tool for fleet risk assessment and prioritization with AI

Outcome

Outcome

Identified friction in clarity, prioritization, and trust

  1. Validated the need for stronger actionability and risk signalling

  2. Defined design evolution toward decision enablement

Identified friction in clarity, prioritization, and trust

  1. Validated the need for stronger actionability and risk signalling

  2. Defined design evolution toward decision enablement

Old Design

Old Design

Users could not decide what action to take next

Pendo showed no progress

Pendo showed no progress

Example of few metrics tracked:

Interviews hit a wall

Interviews hit a wall

Untill..

Untill..

Me: Can you explain what you see on the screen and what you do next?

Interviewee: …..Radio silence….

Then we knew

Then we knew

“they are not understanding the graphs”!

What Changed in Our Approach

What Changed in Our Approach

1

1

AI highlights priority risks within the dashboard

AI highlights priority risks within the dashboard

2

2

Information structured by urgency and impact

Information structured by urgency and impact

3

3

Clear pathways from insight → investigation → action

Clear pathways from insight → investigation → action

4

4

Reduced cognitive effort in identifying “what matters today”

Reduced cognitive effort in identifying “what matters today”

We are evolving SafetyIQ into:

We are evolving SafetyIQ into:

"A decision-intelligence layer that transforms fleet data into prioritized, actionable direction, not just another reporting surface."

"A decision-intelligence layer that transforms fleet data into prioritized, actionable direction, not just another reporting surface."

We won’t be relying on summaries alone

We won’t be relying on summaries alone

Old Program Dashboard

Old Program Dashboard

New Program Dashboard

New Program Dashboard

Usablity Metric Shift

Usablity Metric Shift

Post Launch Value Metrics

Post Launch Value Metrics

We measure if the designs are working and keep improving.

We measure if the designs are working and keep improving.

How did it change Insurance Owner's lives?

How did it change Insurance Owner's lives?

  1. Insurance Program managers spent 15 minutes more time (5% increase) on the platform per week (Data given by roduct oprations)

  2. 30% more users logged into SafetyIQ in 1 month (Pendo Insight). These were the underwriters who were generally just taking information from Program managers.

What success will look like in future?

What success will look like in future?

1

1

Faster risk detection

Faster risk detection

2

2

AI-triggered actions

AI-triggered actions

3

3

Decision confidence

Decision confidence

4

4

Consistent trackable usage

Consistent trackable usage

We are building SafetyIQ into a cognitive layer over fleet data, not just another dashboard.

We are building SafetyIQ into a cognitive layer over fleet data, not just another dashboard.

Thank You

Thank You