Case Study 2023 UX Designer

Statera
Command Center

From passive reporting to active intelligence
for high-velocity ad campaigns

Statera redesigned dashboard with real-time metrics and action center
Scroll to explore
Project Overview

Outline

Ad-Ops managers couldn't detect campaign anomalies before significant budget loss. We set out to transform Statera from a retrospective reporting tool into an active command center, giving teams the intelligence to act before damage was done.

Problem statement

How might we surface campaign anomalies in real time and guide Ad-Ops managers toward the right action, before budget is wasted?

Role UX Designer
Duration 4 months
Team 1 PM, 2 Engineers
Platform Web (Desktop-first SaaS)
Discovery

Journey to problem statement

What started as a routine quarterly business review led us to closely examine the escalation ticket patterns across the Ad-Tech operations domain.

62% Campaign Loss
62% Campaign Loss Escalations Major contributor in ad-ops support tickets
38% Other Tickets (Billing, Access, Integrations, Reporting bugs)

Problem identification methods

Platform usage analytics

6 months of interaction data across 120+ accounts

Ad-Ops manager shadowing

Embedded with 8 teams during live campaign launches

Support ticket analysis

2,400+ tickets analyzed for delayed discovery patterns

Competitive benchmarking

Studied Datadog (severity-based alert triage), Bloomberg Terminal (information density at speed), and PagerDuty (actionable escalation flows)

Insights

Findings

What our users said:

"By the time I see the problem in my weekly review, I've already burned through $15K in wasted spend."

Senior Media Buyer, E-commerce Brand

"The dashboard shows me beautiful charts, but I need it to scream at me when something's wrong."

Performance Marketing Director, SaaS Company
What we analyzed:
  • Ad-Ops workflows are reactive, not proactive. Users logged in 2.1x/day. Critical mid-day issues went unnoticed.
  • Context switching kills efficiency. Managers toggled between 5 platforms 47 times per session.
  • No prioritization framework. All campaigns presented equally. Which one needs attention RIGHT NOW?
Delayed anomaly detection Lack of actionable alerts Fragmented workflow Manual ROI calculations Ambiguous status indicators No campaign prioritization
Platform Limitation

Statera's data pipeline operated on a 4-hour batch processing cycle. For high-velocity campaigns running $500+/hour budgets:

  • A campaign could hemorrhage $2,000 in 4 hours before appearing in the dashboard
  • All analysis happened client-side, so users manually filtered and interpreted data with no intelligence layer
Before
Facebook / Google / TikTok APIs
4-hour batch
Statera Database
Client-side processing
Static Dashboard
After
Facebook / Google / TikTok APIs
Real-time streaming
Intelligence Layer + Alerts
Contextual scoring
Active Command Center
The Solution

From report generator
to mission control
for ad spend.

Framework

Core design principles

01

Alert Before It Hurts

Surface anomalies as they happen, not hours later.

02

Context Over Raw Numbers

Every metric needs a comparison point: vs yesterday, vs target, vs threshold.

03

Reduce to Next Action

Don't make users figure out what to do. Suggest the action.

04

Visual Health Scoring

Status badges (LOSING, SCALING, PROFITABLE) replace ambiguous states.

05

Eliminate Platform Hopping

Embed controls directly in the interface.

Transformation

Dashboard evolution

Before: Passive Reporting
Legacy Statera dashboard with single metric, generic chart, no alerts
  • Single metric card, no alerts
  • Generic chart with dips and spikes that demand manual investigation
After: Active Monitoring
Redesigned Statera dashboard with three hero metrics, action center, and real-time performance chart
  • Three hero metrics with real-time deltas
  • Action Center surfaces problems proactively
Transformation

Prioritized campaign management

Redesigned campaigns view with KPI cards, smart filtering, and health-scored campaign cards Campaigns Command

Top-level KPI cards, smart filtering with automatic triage (Critical tab), and campaign cards with visual health scoring (LOSING, SCALING, PROFITABLE) replace the flat table.

LOSING ROI < 1.0x
SCALING ROI > 2.5x
PROFITABLE ROI 1.2x–2.5x
PAUSED Inactive
Transformation

Integrated campaign builder

Campaign creation wizard with three steps: Basics, Economics, and Guardrails Campaign Wizard

Early testing with a single-page form showed users skipping the guardrails fields (they felt optional). Breaking the flow into 3 explicit steps (Basics, Economics, Guardrails) made each stage feel intentional. Users couldn't accidentally skip the safety net.

1

Basics

Campaign Name + Platform Selection (Facebook, Google, TikTok)

2

Economics

Budget allocation, Target ROI thresholds, Auto-pause triggers

3

Guardrails

Min/max ROI bounds, Daily spend limits, Performance alerts

It's like Statera went from showing me a map
to being my co-pilot.
Ad-Ops Manager, Post-Launch Interview
Outcome

What happened

The redesigned command center was validated through usability testing with Ad-Ops teams and received strong qualitative feedback. However, the company made a strategic decision to sunset Statera before the redesign reached production.

This is the reality of product design: not every well-designed solution ships. The business context shifted, and the project was deprioritized in favor of other initiatives.

What we validated

Real-time alerts changed behavior

In testing, users shifted from reactive weekly reviews to proactive monitoring when anomalies were surfaced immediately.

Health scoring reduced cognitive load

LOSING / SCALING / PROFITABLE badges let users triage campaigns at a glance instead of interpreting raw numbers.

Embedded actions eliminated platform hopping

Users completed campaign adjustments within Statera instead of context-switching across 5 platforms.

Reality

Constraints & trade-offs

4-hour data lag couldn't be fully eliminated

Engineering could reduce the batch cycle to 30 minutes, but true real-time streaming required infrastructure investment the business wasn't ready for. We designed the alert system to work within the 30-minute window and clearly communicated data freshness to users.

Health scoring thresholds were debated

What counts as "LOSING" vs "PROFITABLE" varies by industry and campaign type. We defaulted to ROI-based thresholds but designed the system to let power users customize their own, a scope addition the PM initially pushed back on.

Campaign creation was descoped from V1

The integrated campaign wizard was designed but deprioritized for the initial release. The team decided to ship the monitoring and alerting experience first, since that addressed the core pain point.

Reflection

What I took away

On the Design

Shadowing > Surveys

Watching an Ad-Ops manager manually calculate ROI in a spreadsheet, while sitting in front of our dashboard, was worth more than 50 survey responses. The gap between what users say they do and what they actually do is where the real problems hide.

Status labels are a design decision, not just UI

Choosing to label a campaign "LOSING" instead of showing a red number was a product decision as much as a design one. It required alignment with PM and stakeholders on what the platform should communicate and how directly.

On the Outcome

Good design doesn't guarantee survival

The project was sunset due to business priorities, not design quality. This taught me that understanding the business context (funding, market timing, strategic direction) matters as much as understanding the user.

What I'd push for earlier

I would have advocated for a smaller, shippable slice sooner, just the alert system without the full dashboard overhaul. A faster path to production might have demonstrated value before the business decision was made.