01
Separate website from product
Do not force the marketing site and custom application into the same operating model when their needs differ.
For startups
Use managed hosting for the public website and unmanaged VPS only when the product needs a custom runtime—and when someone on the team owns production operations.
Bitfoo opens for orders soon.
Managed platform
If operations are owned
Avoid speculative infrastructure
Not company age
What matters first
The right early infrastructure is small enough to understand and disciplined enough to hand off.
01
Do not force the marketing site and custom application into the same operating model when their needs differ.
02
A VPS decision requires a named person responsible for security, deployments, monitoring, and recovery.
03
Document code, data, DNS, secrets, and deployment so growth does not depend on one founder’s memory.
04
Change capacity or architecture from observed resource use, reliability needs, or application requirements.
Recommended starting path
The company stage does not decide the product. The runtime and operations model do.
Public website
Use for WordPress, launch content, documentation, and conventional company pages.
Custom application
Use when the product needs persistent services and the team can operate the host.
Relevant services and guides
Give each workload the smallest operating model that satisfies its current requirements.
Company presence
Launch a managed company site without coupling it to the product’s server stack.
Explore shared hosting →Custom product
Run a custom application when the team accepts the unmanaged server scope.
Explore VPS →Workload choice
Review Next.js, Rails, Docker, Laravel, Node.js, Python, and Django guidance.
Browse workloads →Before committing
The infrastructure should be understandable by someone who did not participate in every early decision.
Record domains, DNS, hosting, repositories, secrets, data stores, integrations, billing, and renewal owners.
Make the marketing site and product release paths repeatable and separate.
Observe the core customer path and dependencies before adding infrastructure for hypothetical scale.
Restore or redeploy to a clean environment so the company is not dependent on one machine or founder.
Use shared hosting for a conventional company website. A custom product may need a different platform or VPS from the beginning, depending on its runtime and the team’s operational capabilities.
Yes. They often should. Separate them when their deployment, availability, security, or ownership requirements differ.
Move a workload when it requires root access, custom services, containers, persistent processes, or server-level control—and when someone can operate the server.
Not by itself. A larger profile adds capacity. Application health, releases, monitoring, backups, recovery, and multi-host architecture remain separate decisions.