How would you work with engineering to reduce latency in Azure

Problem Statement Description

Product context: Microsoft is a productivity, software, AI, gaming, and cloud company; its products include Windows, Microsoft 365, Teams, LinkedIn, Xbox, Azure, Dynamics, and Copilot. Azure is Microsoft's cloud platform; its products include compute, storage, databases, networking, identity, AI services, analytics, and developer tools.

You are a Technical PM on Microsoft Azure focused on SMB customers who rely on Azure for business-critical workloads but may not have large infrastructure teams to diagnose performance issues. Customers are reporting that parts of their Azure experience feel slow, which could include portal interactions, API calls, resource provisioning, data access, or application response times for services they run on Azure.

Your task is to explain how you would work with engineering to reduce latency in Azure. The discussion should cover how you would clarify the problem, define what “latency” means in this context, identify affected users and workflows, prioritize investigation areas, and partner with engineering through diagnosis, implementation, rollout, and measurement.

This is a technical PM problem, so focus on how you would bridge customer impact, product requirements, system constraints, and engineering execution. Avoid jumping directly to a technical fix; instead, show how you would structure the work, make trade-offs, and ensure improvements are safe, measurable, and aligned with Azure’s reliability, security, and enterprise-trust expectations.

The experience should consider:

- Which Azure workflows, services, regions, APIs, or customer segments are in scope, especially for SMB customers

- How latency should be defined and measured, including client-side, server-side, network, API, and end-to-end user-perceived latency

- What data and instrumentation would be needed to segment the issue by geography, workload type, SKU, tenant size, time of day, and service dependency

- How to distinguish product latency from customer architecture, network conditions, third-party dependencies, or misconfiguration

- How you would work with engineering on requirements, root-cause analysis, prioritization, trade-offs, and validation

- Reliability, availability, privacy, security, compliance, and backward-compatibility constraints in a cloud platform environment

- Rollout planning, experimentation, observability, incident readiness, and rollback criteria

- How success would be measured through customer-facing metrics, engineering health metrics, and business impact

The goal is to demonstrate how you would lead a cross-functional, technically grounded effort to improve Azure latency while protecting platform reliability and customer trust.