SaaS vs Self-Hosting: When the Invoice Lies

Open source is not free. You trade invoice cost for attention: backups, upgrades, monitoring, and overnight failures. That is true for a VPS and for “self-host on a cloud account” (Workers, R2, Durable Objects). A useful comparison is total cost of ownership, not license vs license.

Ankik beta (~50 users), July 2026: AWS EC2 t3.small ($18) + RDS db.t3.micro ($16) + EBS/transfer (~$5) ≈ $39/mo (~Rs 3300), USD billing, ~180ms from Gujarat in that placement. Same workload moved 10 July 2026 to a private Mumbai Linux VM (2 vCPU, 4GB, self-hosted Postgres) at Rs 520/mo, INR/UPI, ~28ms from Vadodara. About 84% lower infrastructure bill at those snapshot prices—excluding the time to operate the VM, backups, and incidents.

Business fit: does the alternative cover the actual job? Team capability: who patches, restores, and owns security updates? Self-hosting cost: VM or usage quotas, storage, egress, and hours per month—not license alone. Security and compliance: data location, access control, backup encryption. Migration risk: export path, downtime budget, rollback.

Self-host usually wins with steady load, a region/billing model that fits (for example India-first products that care about Mumbai latency and INR billing), size that fits a well-sized VM or free-tier quotas, and at least one person who will run backups and restore drills. Stay on managed SaaS when no one will own ops; you need point-in-time recovery, multi-region, or an uptime story you cannot staff; or the math is weak. We often tell teams on Supabase Cloud ~$25 to stay put until usage justifies the ops tax.

Cancel a vendor only after the five-point check and a before/after cost sheet that includes ops time. If that list feels heavy, keep RDS, Supabase Cloud, or managed mail. Reframe ($199) is diagnosis only so the recommendation can be Keep.

Evidence Sources & Reference Slugs