🛠Death to the backend
Investing in developer tools that give frontend developers superpowers.
Even in 2022, too much time is wasted writing boilerplate code. The way developers build software has improved dramatically, but much work still needs to be done. The past decade has been focused on abstracting away complex physical infrastructure. I believe that the next decade will be focused on abstracting away the role of the backend developer.
I want to invest in developer tools that give frontend developers superpowers. These tools are adopted in the early days, making it faster for frontend developers to build working products and go from 0 to 1. Once integrated, these tools stick around for a long time and accumulate customer data, becoming deeply entrenched and painful to migrate off of. And like other developer tools businesses, they're often high margin and see revenue grow with their customers' success.
In Database in The Browser, Stepan Parunashvili (cofounder of Instant) declares, "Much of backend development ends up being a sort of glue between the database and the frontend." Instant (YC S22) gives frontend engineers superpowers. They're building a synchronized, real-time database that frontend developers can query directly, bypassing the need for a cumbersome backend. This type of database powers multiplayer applications we all love, like Figma, Linear, and Notion. In the age of remote collaboration, it's clear that more developers will want to add features like this to their products, and Instant offers the path of least resistance.
I find Backend-as-a-Service companies compelling because of their ability to be used for many different use cases. Startups like Prisma (Series B) and Supabase (Series B) are building horizontal platforms and can be implemented alongside existing backend code. At the same time, there's an opportunity for more specialized approaches that can better serve major use cases.
Headless content management system (CMS) platforms are BaaS that focus exclusively on powering the data layer for content applications. These are features as small as product documentation or as large as multinational news homepages. Some early companies in this category are Payload (Seed), Strapi (Series B), and Sanity (Series B).
The "headless" model applies to areas beyond CMS, such as headless commerce platforms. Chord (Series A) provides the rails for direct-to-consumer brands to build unique online stores without reimplementing core e-commerce functions. I believe commerce platforms are more substantial businesses than CMS platforms because their pricing is generally transaction based, as opposed to the traditional SaaS model.
There are opportunities for new headless categories to emerge, like how Tailor (YC S22) is building headless ERP for Japanese enterprises. Investing in new headless categories may offer a chance to back an early player in a nascent market.
Most of my career has been at developer tools companies such as Twilio and Mux. As a backend developer, I've spent thousands of hours connecting different APIs and gluing together databases. I've seen first-hand how backend engineers quickly become an expensive bottleneck in the modern software development lifecycle. We are in the early innings of the software revolution, and the tools to support software developers will continue to evolve as the industry grows.
Seems like they are so many opportunities with develop tools. Can you talk more about what it’s such a challenge to build? Not that I’ve built one myself, but what’s the biggest issues? Are our backend engineers not trained to be innovative enough?