We pay Heroku many, many tens of thousands of dollars a year. And I still use free dynos, both personally and at work. For example, throwing up a quick app for testing (where I'm happy with dynos that sleep or are limited per-repo). By pushing us off the Heroku ecosystem for some stuff, we might as well just move everything.
The only reason we even use Heroku now is because I used it for free over a decade ago.
I get why they made this decision, and I'm excited for Fly.io, Render, etc who can run the same playbook Heroku did 15 years ago. But also a bit sad, from a nostalgic standpoint. Many of us are here because of Heroku's free tier, and I'm very thankful for it.
When I looked at Heroku pricing way back when, I immediately had a "WOW that's expensive" reaction.
I look at it now, and...well, I'm all-in on k8s for most things, and cloud functions for most everything else, so I'm really not sure what the advantage of using Heroku would ever be if it they don't have a free plan.
For free databases there are multiple options like CockroachDB and Supabase; throw up a $6/month droplet at DigitalOcean and you get the equivalent of a $50/month dyno at Heroku. Yes it's easier to deploy to Heroku, but it's only a couple hours to set up some kind of CI/CD deploy, and then you can control it more precisely.
Heroku has basically been a "first one is free, but as soon as the business gets big, soak them" company from the start. Given the number of companies offering free levels of cloud functions and hosting, I think that's where most new experimental development will migrate to in the future.
I sympathize with them for giving up in the fight against abuse of their free services, but ... well, I think they're likely to transition to irrelevance if they don't pivot or slash prices soon.
I agree that paid options of heroku hasn't been competitive for a long time. But removing the free, hobby tier suddenly makes hundreds of programming tutorials outdated which affects students and job seekers, I highly doubt that the students from obscure schools from around the world would be covered by the to-be announced education program and these are the groups who cannot spend any money.
Fortunately there are great alternatives[1] for PaaS and for those who can manage the server there's Oracle Cloud free-tier which offers 24GB RAM 4 Core ARM Server which can be used as single machine or split into 4 VMs and 2 x86 1GB RAM 2 Core VMs with 200 GB Block Volume in total. But Oracle Cloud free-tier has lots of quirks and often requires a script[2] to provision the machines(Due to demand).
Yeah, I really wonder just how out of touch they are with their customer base to believe this is the best course of action. So many of their paying customers ended up that way because it was free and easy to test out an idea. Then if you started making money it was still easy and you could afford to pay Heroku to scale. Then a lot of these paying customers will never outgrow Heroku scale so they stay forever.
I got the same sentiment. Heroku for us was number one because of the free tier.
They could consider keeping an option to have 1-2 free instances for already paying customers. It also feels to me that their business will slowly fade since now.
No, this is about optimizing profitability. They were bought by Salesforce and that whole cohort (Salesforce, Oracle, et al) are machines optimized to extract as much money as possible from the Fortune 1000.
Salesforce doesn't need a free tier to generate product leads. They have an army of sales people to push the product to customers that will write them big fat checks, then hand down edicts from the CTO's office requiring their internal devs use the technology.
yup, and their enterprise clients don't care, maybe they even prefer that. in the mean time, it gets salesforce enough capital to leverage to acquire the next generation of leaders to stagnate and extract. the wheel spins on...
I don't know if I would classify it as an unrecoverable death spiral, but it does seem very short-sighted and ignorant of their customer base.
Heroku has two core tenets - developer experience and upgrade to paid tier for production workloads.
As a result, at least the Heroku customers I know use the free tier for a good portion of their _non_ customer-facing production workloads - prototypes, staging, administrative functions, and so on.
This both increases their cost and makes budget planning a lot less stable. It means developers may become motivated to start coming up with workflows that target other environments where they would normally target Heroku, which like robs Heroku of a source for future revenue. Once you take on the devops work yourself, Heroku is no longer price competitive.
In other words, this new cost makes the unique value of Heroku look quite a bit more like a detriment. Thats rather unfortunate.
But Heroku also had a major GitHub integration that devolved into CVE which could not be fixed for about a month (among other bad news parts of that story, such as lying about the scope and slow walking the disclosures, that all seemed to just get worse.)
This is ostensibly all the result of the brain drain after SalesForce acquisition has set in. It's a death spiral.
Nah this is a bean counter decision, not an engineering or marketing decision. It will be bad mojo. There are a dozen other ways to cut down on the bogus/abusive accounts.
Why not bare metal servers or VPS from like... DigitalOcean? What is DigitalOcean missing for your use case/preference specifically compared to Next + Vercel?
I'm not a DigitalOcean shill or employee, I'm just curious what I'm missing from a "what do other competitors out there offer".
I always thought it was like... spinup Debian/Ubuntu VPS, ssh to it, install Docker, run docker-compose or Docker Swarm or... Terraform?
If you scroll down a little there's a section titled "Out-of-the-box features" that answers your question. I think the edge functions would be the hardest thing to do on your own.
For just the hosting, maybe. But there's so much more Heroku does, from spinning up test environments for PRs to storing secret keys (across different repos) to being a CI to monitoring to... so so much more.
Hooking up some kind of CI/CD to GitHub through webhooks
> storing secret keys
Built into GitHub (or an instance of Hashicorp Vault which can be hosted for free)
> to monitoring
Can run your own Grafana/Promtheus
Obviously there's a cost running all of this yourself as opposed to just paying them to do it. Just making sure I wasn't missing something obvious tradeoff wise between "our company would rather pay somebody to manage all of this for us"
> Services like Herokus are useful because they save hundreds of hours of sysops.
Which they know, and can charge you accordingly for, right?
If a single sysop engineer cost $100k/yr (without anybody managing them), they can charge you $50k/yr to replace them and it'd still be a steal, right?
Assuming they have no competition, yes. But now they're not just competing with sysops, they're also completing with a bunch of other app platforms. None of the others are quite as full featured, but if the competitors can save 80% of the sysops time at 20% of the cost, Heroku will see people switching away.
As someone who does what you suggest, it's great but it is a lot of overhead and not zero click. Updates, reboots, lambda functionality (autoscale, blue/green, etc) and database hosting is always complex.
The only reason we even use Heroku now is because I used it for free over a decade ago.
I get why they made this decision, and I'm excited for Fly.io, Render, etc who can run the same playbook Heroku did 15 years ago. But also a bit sad, from a nostalgic standpoint. Many of us are here because of Heroku's free tier, and I'm very thankful for it.