Skip to content
Module
Tech

What are microservices? Definition and trade-offs

A microservices architecture splits a product into small autonomous services, each handling one business capability and deployed on its own. Services communicate over the network, through APIs or messages, and each one ideally manages its own data. A team can then change and ship its service without waiting for the others.

Definition written by Alex Dionisio, Senior Product Manager. Updated November 13, 2026.

Why it matters for a PM

Microservices promise team autonomy and the ability to scale one part of the product without touching the rest. In exchange, they add network latency, partial failures, distributed data and a strong need for monitoring tools. For a PM, this changes how work is split, how a feature spanning several services is estimated and how an incident is diagnosed.

Example

On a food delivery platform, a new group order option touches the cart, payment and delivery tracking, three services owned by three teams. The PM sets up a joint review of the API contract between the services and specifies the expected behavior when the payment service responds slowly.

Key points

  • Each service deploys independently, which supports team autonomy.
  • A cross-cutting feature requires coordinating several services and their contracts.
  • Network calls between services add latency and partial failure cases.
  • Without central logs and traces, finding where an error started becomes hard.

Common mistakes

  • Splitting into microservices with a small team and a product still searching for its market.
  • Sharing one database across services, which cancels out their independence.
  • Underestimating the cost of running and monitoring dozens of services.

Go further with Module

The courses and lessons that cover this concept:

Guides on this topic