Ask HN: Why are people paying so much for Vercel?

Hacker News Discussion ↗

Why people pay for Vercel

  • Main reason: convenience and avoiding DevOps / infra work. “Run vercel and it just works” for many frontend/Next.js projects.
  • Handles CI/CD, preview deployments per PR, global CDN, caching, edge functions, and Next.js-specific complexity (e.g., version skew with React Server Components).
  • Especially attractive to:
    • Small startups where engineering time is the scarcest resource.
    • Agencies and freelancers maintaining many small/medium client sites.
    • Teams focused on UX/product, not routing policies, IAM, servers, etc.
  • Dynamic scalability and not owning physical infrastructure are seen as key benefits, similar to AWS Lambda.

Cost vs DIY / Alternatives

  • Many argue Vercel’s markup over raw AWS/VPS/bare-metal is high and unnecessary for low-traffic apps.
  • Alternatives mentioned: AWS (Lambda, Amplify), Google Cloud Run, Azure Container Apps, Hetzner, Cloudflare Pages/Workers, SST, Supabase, Dokku, Coolify, Docker Compose, self-managed Kubernetes.
  • Some say a few hours with Ansible + a cheap VPS can get you production-ready for a fraction of the monthly cost.
  • Counterpoint: infra setup, security, backups, and ongoing patching/monitoring can easily cost more in dev time than a $20–$100+ monthly Vercel bill.

Scaling, lock‑in, and “premature optimization”

  • Common stance: use Vercel early; only move off once bills hit high hundreds/thousands or you hit resource limits.
  • Others argue you should learn infra early to avoid future re‑architecture and vendor lock‑in.
  • Some see Vercel as solving complexity it helped create (Next.js + RSC), making “growing up” to other platforms harder.

Skills, culture, and attitudes

  • Debate over whether developers are increasingly “infra-illiterate” and overly reliant on abstractions.
  • Some insist basic Linux/VPS skills are easy (especially with modern tooling/AI help); others say it’s a steep hill for pure frontend devs.
  • Several note that most products never reach real scale, so heavy infra investment is often wasted.

Criticisms and concerns

  • Complaints about high enterprise pricing (e.g., SAML), with reports of quick, cheap migrations to competing services.
  • Worries about brand damage from scammy sites on the free tier.
  • Perception that marketing, control of Next.js, and influencer sponsorships drive adoption as much as technical merit.