General Merchandise Retailers2021Machine Learning (classification)Predictive AnalyticsB2C
Marks & Spencer

Marks & Spencer reduces MTTR by one-third and ends 2 a.m. incident rooms with New Relic

After shifting to digital-first retail in 2018, Marks & Spencer deployed New Relic to gain end-to-end observability, reducing MTTR by one-third, cutting incident team sizes from 80 people to a targeted notification, and achieving full service visibility across thousands of digital touchpoints.

MTTR Reduction33% (~1/3)
5 min read

Background

M&S's digital transformation introduced hundreds of microservices and multicloud deployments that classical monitoring could not track holistically. Incident response was slow, expensive, and team-intensive.

What Was Implemented

  • Deployed New Relic observability across multicloud and microservices environments (Azure, AWS, Kubernetes)
  • Unified logs, events, metrics, and traces into a single observability layer
  • Built refined dashboards incorporating context, analysis, and business-level aggregation
  • Applied AIOps capabilities to reduce false positives and contextualize anomalies

Results

M&S reduced MTTR by approximately one-third (confirmed). Before New Relic, incidents involved up to 80 people; after, the platform pinpoints the issue and notifies only the responsible team, enabling resolution in minutes. The 80% incident-detection-time improvement cited in the book was not found in the New Relic source article.

Lessons

  • Unifying observability data into a single platform is a prerequisite for reducing incident team size from "everyone" to "the right team"
  • Omnichannel retailers face unique observability challenges because customer journeys span mobile, web, and in-store systems simultaneously
  • Moving from averages and percentiles to context-aware anomaly detection gives engineering teams the precision needed to protect high-value customer cohorts

Ready to implement AI in your commerce operations?

McFadyen Digital helps teams move from case study to live implementation.

Talk to an expert →