On February 6th, 2026, Salesforce quietly dropped a bombshell: Heroku, the platform that defined modern cloud deployment for an entire generation of developers, is entering "sustaining engineering mode." No new enterprise sales. No new features. Just security patches and keeping the lights on.
For a platform that once made git push heroku main feel like magic, this isn't a sudden death — it's the final stage of a slow decline that started years ago. And it signals something bigger happening across the entire Platform-as-a-Service landscape.
As a team that builds and operates hosting infrastructure at Sid Techno, we've been watching this unfold closely. Here's our honest take on what happened, what it means, and where the industry is headed.
What Actually Happened to Heroku
Salesforce acquired Heroku back in 2010 for $212 million. For years, Heroku remained the gold standard for developer experience — spin up a dyno, push your code, and you're live. It was beautiful in its simplicity.
But the cracks started showing long before 2026:
- November 2022: Heroku killed its free tier, pushing hobby developers and students off the platform overnight. This was the first mass exodus.
- 2023-2025: Feature stagnation set in. While competitors shipped container orchestration, edge functions, and GPU support, Heroku's roadmap went quiet. Salesforce was funneling engineering resources elsewhere.
- February 2026: The official announcement — sustaining engineering mode. Enterprise contracts frozen for new customers. Existing customers can renew, but the writing is on the wall.
The root cause? Salesforce's strategic pivot to AI. When your parent company is betting everything on Agentforce and enterprise AI, a developer platform that serves startups and indie hackers doesn't make the priority list. Heroku wasn't failing — it was deprioritized.
The Broader PaaS Market Is Restructuring
Heroku's decline isn't happening in isolation. The entire PaaS landscape is going through a fundamental restructuring.
Netlify Is Shrinking
Netlify, which once seemed poised to own the Jamstack ecosystem, has gone through multiple rounds of layoffs since 2022. The company that pioneered static site deployment has struggled to expand beyond its original niche as the market shifted toward full-stack frameworks and server-side rendering.
Vercel Is Growing — But Narrowing
Vercel has captured significant market share, largely riding the success of Next.js. But their growth is tightly coupled to a single framework ecosystem. If you're not building with Next.js, Vercel's value proposition weakens considerably. That's a strategic risk the market hasn't fully priced in yet.
The New Guard Is Rising
The real winners in this shakeup are the platforms that learned from Heroku's simplicity while addressing its limitations:
- Railway has arguably the smoothest migration path from Heroku. Similar concepts, better pricing, superior developer experience. For most developers building new projects, Railway hits the sweet spot between ease of use and capability.
- Render mirrors Heroku's mental model almost exactly — web services, background workers, managed Postgres, cron jobs — making the migration path straightforward. Their free tier also makes it a natural landing spot for developers who were pushed off Heroku's free plan.
- Fly.io takes a different approach with lightweight VMs across 30+ regions. If latency matters for your application, Fly.io's global deployment model is compelling.
- DigitalOcean App Platform offers a good middle ground between raw infrastructure and fully managed PaaS, backed by a company with strong financial fundamentals.
What This Means for Developers
If you're currently running production workloads on Heroku, don't panic — but do start planning. Here's a realistic assessment:
Short Term (Next 6 Months)
Heroku isn't shutting down tomorrow. Your apps will keep running. Existing enterprise contracts will be honored. But you should start evaluating alternatives now while you have the luxury of time rather than being forced into an emergency migration later.
Medium Term (6-18 Months)
Expect Heroku's ecosystem to thin out. Third-party add-on providers may start deprioritizing Heroku integrations. Community support will dwindle as developers migrate away. The buildpack ecosystem will stagnate. This is where the real pain starts — not from Heroku breaking, but from the ecosystem around it slowly eroding.
Long Term (18+ Months)
At some point, Salesforce will need to make a call on Heroku's future. Sustaining engineering mode is rarely a permanent state — it's usually a transition toward either a sale or a sunset. Plan accordingly.
Choosing Your Next Platform
Here's our honest framework for evaluating where to go next. There's no single right answer — it depends on what you're building:
For Startups and Side Projects
Railway or Render are your best bets. Both offer Heroku-like simplicity with modern infrastructure underneath. Railway edges ahead on developer experience; Render wins on the free tier for prototyping.
For Performance-Critical Applications
Fly.io if you need global distribution with low latency. Their VM-based approach gives you more control over your runtime environment than container-based platforms.
For Teams That Want More Control
Consider platforms like Coolify (self-hosted PaaS on your own servers) or managed Kubernetes options. You trade simplicity for control and cost savings at scale.
For WordPress, PHP, and Laravel Workloads
This is an area we know well. Our own platform, DeployBase, handles WordPress, Node.js, PHP, and Laravel hosting on dedicated Hetzner infrastructure. We're not trying to be the next Heroku — we focus specifically on making deployment simple for these specific stacks. It's one option among many, and it's honest to say we're still a smaller player compared to Railway or Render. But if your workload fits our stack, it's worth a look.
For Enterprise Workloads
AWS, GCP, or Azure remain the safe choice for large-scale enterprise deployments. The major cloud providers aren't going anywhere, and their PaaS offerings (Elastic Beanstalk, App Engine, Azure App Service) are battle-tested, even if they lack the developer experience polish of newer platforms.
Lessons for the SaaS Industry
Heroku's trajectory holds important lessons for anyone building or running SaaS products:
1. Acquisition Can Kill Innovation
Heroku was a beloved, innovative product before the Salesforce acquisition. Under corporate ownership, it became a line item on a balance sheet. When the parent company's strategy shifts, acquired products get caught in the crossfire. This is a cautionary tale for every startup considering an acquisition offer.
2. Developer Trust Is Earned Slowly and Lost Quickly
Killing the free tier in 2022 was the beginning of the end for Heroku's developer mindshare. Developers choose platforms for personal projects first, then bring those platforms into their companies. When you cut off the grassroots pipeline, the enterprise pipeline eventually dries up too.
3. Platforms Must Keep Evolving
The cloud infrastructure space moves fast. Containers, edge computing, serverless functions, GPU workloads — the platforms that survive are the ones that keep shipping. Heroku stopped shipping, and the market moved on.
4. Diversification Matters
Vercel's tight coupling to Next.js is a risk. Netlify's bet on Jamstack proved too narrow. The platforms showing the most resilience are the ones serving multiple use cases and technology stacks.
Our Perspective as Builders
At Sid Techno, we build and operate multiple SaaS products — from hosting platforms to help desk software to logistics systems. We run everything on dedicated Hetzner infrastructure, and we've experienced firsthand the tradeoffs between managed PaaS platforms and running your own infrastructure.
The Heroku situation reinforces something we believe deeply: owning your deployment stack matters. Not in a "build everything from scratch" way, but in a "don't build your house on someone else's land" way. Whether that means using a transparent, well-funded platform or investing in self-hosted tooling, reducing single-vendor dependency should be part of every technical strategy.
The PaaS market isn't dying — it's evolving. The simplicity that Heroku pioneered is now table stakes. The next generation of platforms will compete on pricing transparency, regional availability, AI-native tooling, and ecosystem breadth. That's good for developers.
Key Takeaways
- Heroku isn't dead yet, but it's on life support. Sustaining mode means no new features, and no new enterprise sales. Plan your migration timeline now.
- The PaaS market is consolidating. Smaller platforms will either grow or get acquired. Railway and Render are the current frontrunners for Heroku refugees.
- Don't rush your migration. Evaluate platforms based on your actual workload, not hype. A bad migration is worse than staying on a stable-but-stagnant platform.
- Reduce vendor lock-in. Use containers, standardize your CI/CD pipelines, and make sure you can deploy your app to multiple targets. The next platform shift is always coming.
- Watch Salesforce's next move. Sustaining mode is usually temporary. Whether Heroku gets sold, sunset, or somehow revived will shape the market for years to come.