They promised the cloud would be cheaper, faster, and easier.
For us, the savings never materialized and the team never shrunk.
We left the cloud
After using clouds from both Amazon and Google extensively over the past 15 years, we finally had enough of the outrageous bills and the ever-increasing complexity. So in 2023, we pulled Basecamp, HEY, and five other heritage apps out of AWS and onto our own hardware — without adding any new staff.
The bottomline:
Leaving the cloud will save us
~$10 million over five years.
That's a reduction in our infrastructure costs of between half and two-thirds.
Our cloud exit in full detail
- Why we’re leaving the cloud
- Five values guiding our cloud exit
- The hardware we need for our cloud exit has arrived
- Cloud exit pays off in performance too
- We have left the cloud
- The Big Cloud FAQ
- We stand to save $7m over five years from our cloud exit
- Our cloud exit has already yielded $1m/year in savings
- It's five grand a day to miss our S3 exit
- Our cloud-exit savings will now top ten million over five years
- Introducing Kamal
- Kamal 1.0
- Kamal 2: Thou need not PaaS
- Our switch to Kamal is complete
- Keeping the lights on while leaving the cloud
- De-cloud and de-k8s — bringing our apps back home
- Servers can last a long time
- Don’t be fooled by serverless
- The only thing worse than cloud pricing is the enterprisey alternatives
- Sovereign clouds
Our cloud exit made headlines
- The Register: Save $7 million on cloud by spending $600k on servers, says 37signals’ David Heinemeier Hansson
- Techspot: Basecamp decided to leave the cloud after spending $3.2 million in one year
- TechCrunch: Could machine learning refresh the cloud debate?
- Data Centre Dynamics: 37signals expects to save $7m over five years after moving off of the cloud
- Network World: Cloud vs on-prem: SaaS vendor 37signals bails out of the public cloud
- Banerjee: Leaving the Cloud
- IT Web: Why cloud isn’t always the answer