BFF - Backend for Frontend - Pattern in Microservices

Watch the video explanation ➔

Say, we are launching an e-commerce website that people can use from their desktop and place orders. We want to render the product details that include - name, description, sellers, variants, reviews, and FAQs. The site did well, and now we decide to launch a mobile app.

Given that the mobile has limited real estate, it is very hard to render the same information that we do on the Web. Hence, we choose not to render Reviews and FAQs. Thus, even if we receive the reviews and FAQs in the response, while rendering we choose not to render them.

This is a waste of user bandwidth, data, and processing power. Ideally, we should not be sending unnecessary fields from the backend. So, how do we implement this?

Backend for Frontend

BFF is a layer that sits between the clients and the backend. Every type of client has a dedicated BFF; for example, Desktop Web has its own BFF, while Mobile has its own.

Depending on the request, the BFF then talks to the backend, grabs the data, filters out the unnecessary fields, and responds. It can also optionally transform the data in a client-specific format.

This way, we keep the backend simple and apply all presentation-level hacks and tweaks to BFF.

BFF and Microservices

BFF acts as a perfect abstraction for underlying microservices. For an API, it can connect to necessary microservices, gather the responses, and respond. This ensures that we are not fetching data that we don’t need.

For example, Given compatibility and constraints, a Desktop BFF can talk to Orders, Sellers, and Reviews Services, while a Mobile BFF can talk to Orders, Sellers, and AR services to respond to the same API endpoint.


  • we can add client-specific tweaks and hacks on BFF
  • we can hide sensitive information from specific clients
  • we can patch client-specific vulnerabilities on respective BFF
  • we can pick the best communication stack for the client and BFFs
  • we can have a general-purpose backend supporting all kinds of clients


  • a large chunk of code would be duplicated
  • needs to support high fan-out, hence pick a stack that suits
  • there is a slight increase in latency with the new network hop
  • by adding BFFs we are adding more moving parts that need to be managed, maintained, monitored, and deployed

Adopting BFF

We should adopt BFF when

  • the interface between different clients varies significantly
  • the communication format/protocol is different from what your backend supports (e.g.: legacy integration might need XML while your backend serves JSON)

Here's the video ⤵

Courses I teach

Alongside my daily work, I also teach some highly practical courses, with a no-fluff no-nonsense approach, that are designed to spark engineering curiosity and help you ace your career.

System Design Masterclass

A no-fluff masterclass that helps experienced engineers form the right intuition to design and implement highly scalable, fault-tolerant, extensible, and available systems.

Details →

System Design for Beginners

An in-depth and self-paced course for absolute beginners to become great at designing and implementing scalable, available, and extensible systems.

Details →

Redis Internals

A self-paced and hands-on course covering Redis internals - data structures, algorithms, and some core features by re-implementing them in Go.

Details →

Writings and Learnings

Knowledge Base



Arpit's Newsletter read by 90000+ engineers

Weekly essays on real-world system design, distributed systems, or a deep dive into some super-clever algorithm.