Securitas AB
Callouts Insight: turning raw data into operational decisions
Ops teams could see raw callout totals but couldn't tell where to act. I designed the KPI and reporting layer that turned fragmented data into a decision-ready view across 50+ markets.
- Role
- Senior UX Designer, Product & Strategy
- When
- Aug 2024 – Mar 2026
- Also on Behance
- Original deck ↗
Want to ask about this project?
AI answers grounded in Anshula's CV and portfolio.
Securitas is one of the world's largest private security companies, providing guarding, alarm response, and mobile services across 50+ markets. Clients needed more than raw data. They needed actionable insight into when, why, and how security incidents were managed.
The problem, and what success needed to mean
Clients were asking questions like "How many callouts did we have this month?" and "Which sites are high-risk?" with no fast way to answer them. Callout data existed in reports, but surfacing a trend meant manually digging through them; there was no instant way to assess SLA compliance or catch recurring issues at a glance.
Before designing anything, I worked with the ops team to define what "solved" would actually mean: a manager should be able to tell whether things were improving or not without opening a spreadsheet, and catch an SLA risk before a client did. That target shaped the decisions that followed.
The constraint
The callout data lived in fragmented legacy reporting systems across 50+ markets, with no shared schema and no consistent definition of what counted as a callout versus an incident. I had roughly three months to design something that worked within that reality, not a rebuilt data layer, while still giving managers a view they could trust.
The tension, and the call I made
Client and ops leadership initially wanted a single comprehensive dashboard that matched the full scope they'd been promised early on. I proposed a phased release instead: a high-level KPI view first, validated with real usage inside the three-month window, then detailed drill-downs in phase two.
The case I made was that a fast, validated first version would surface problems, like how to handle cancelled callouts fairly, while there was still time to fix them, rather than finding out after a bigger rollout.
Approach
I led the research and design end to end: workshops with client operations and dispatch teams, analysis of feedback on the existing summary dashboard, and a full audit of the callout lifecycle, done in partnership with business analysts to map where SLA enforcement and visibility were breaking down.
What shipped, tied to outcome
- Performance indicator cards with trend deltas → cut analysis time 40%, since managers no longer manually compared week-over-week numbers themselves.
- Response-time buckets (under 30 min, 30–60, over 60) benchmarked against SLA thresholds → drove the 25% faster catch on delayed responses.
- Cancelled callouts kept visible for cost tracking but excluded from performance averages → gave regional managers a clean, trustworthy signal without hiding cost data from finance.
- Rollout across 50+ markets, adopted in 200+ client review sessions → became the shared reference point client-facing teams used to defend SLA commitments.
What I'd revisit
Post-rollout, we found blind spots the initial model hadn't accounted for: low-volume sites and missed escalations weren't well served by the summary view, and catching that required a validation loop with regional teams after launch rather than solving it upfront.
At a glance
- Reduced Analysis Time by
- ↓0%
- Improved Alert Response by
- ↑0%
- Markets Served
- ↑0+
Want to talk through this kind of work? Let's talk.
More work
