DEV Community

Yerkebulan Rakhimov
Yerkebulan Rakhimov

Posted on

5-year cost of my .tech domain: $488 vs $204. The renewal trap I missed.

Last week I shipped TubeMine, a free YouTube comments to CSV extractor, in about 2.5 hours of focused work. Phase 0 prototype, validating whether anyone wants the tool before investing more time.

Part of the launch checklist was buying a domain. I bought tubemine.tech through the local Kazakh registrar ps.kz at $8 for year 1. Vercel had $7.99. One cent. Not so much.

This post is about why that cent will cost me $284 over 5 years, and the decision framework I'm taking from it.

The renewal trap

Three days after launch, a reader commented under my build-in-public thread:

Did you check the renewal price? Local registrars usually have it around 20x higher than the promo registration.

I hadn't. Most checkouts only show year-one pricing, with renewal hidden somewhere in the management UI you can't see until after purchase.

The renewal price for ps.kz is buried in a sub-tab called "Финансы" inside the already-purchased domain card. Year 2 onward: 55 800 ₸ (about $120 USD).

For comparison, here is how .tech pricing breaks down across registrars I checked:

Registrar Year 1 (promo) Year 2+ renewal
ps.kz $8 $120
Vercel Domains $7.99 $49
Cloudflare Registrar ~$48 (no promo) ~$48

5-year total cost: ps.kz $488 vs Vercel $204 vs Cloudflare $240.

The $284 delta between ps.kz and Vercel is pure markup on the wholesale renewal price. Identity Digital, which operates the .tech registry, charges $48-49 per year wholesale to anyone. Local registrars stack about $72 per year on top, forever.

Beyond price: friction my $0.01 saving did not cover

The cent isn't where it ended. With ps.kz I also got:

  • DNS activation took 3 days. Their support works business hours only, no weekends, no 24/7. Filed a ticket Saturday, got a response Monday afternoon.
  • Manual DNS setup after activation. No auto-propagation to the Vercel hosting that needed it.
  • Renewal price invisible until after purchase, as mentioned above.

Vercel Domains pushes DNS within 60 seconds because it's the same dashboard as the hosting. Cloudflare does instant propagation across their anycast network.

The decision framework I'm taking forward

This isn't "local registrars bad, Western registrars good." It's a risk-stage tradeoff:

phase: validation (likely to die in 30 days)
  cheap y1-promo registrar OK
  if product dies in month 1, you lose $8 and walk away

phase: validated (signals fired, project surviving)
  transfer to Vercel or Cloudflare at month 11 of year 1
  transfer fee ~$8, adds +1 year to term, skips the markup forever

phase: brand-critical (paying customers, public launch tied to domain)
  Cloudflare or Vercel from day 0
  ~$40 y1 premium buys 24/7 support and no DNS surprises
Enter fullscreen mode Exit fullscreen mode

tubemine.tech is currently in the validation phase. The product got its first paying $19/mo customer earlier today, which puts TubeMine on the validated track. I'll transfer the domain to Vercel at month 11.

What I'm changing operationally

  • Auto-renewal off on the ps.kz domain. Don't want a $120 surprise next year.
  • Calendar reminder at month 11 to evaluate: if TubeMine is still alive, transfer to Vercel; if not, let the domain expire.
  • For future projects: I read the renewal column before the registration column. Always.

The takeaway

The visible price during checkout is rarely the price you pay. Especially for vanity TLDs like .tech, .io, .ai, the wholesale renewal price set by the registry is the floor, and registrars stack their markup on top.

If you're shipping a side project this week, the cheap registrar is fine. If you're shipping anything you expect to live past year 1, check the renewal price BEFORE checkout, not after.

Where do you register domains for new SaaS or side projects? Cloudflare or Vercel straight away, or do you also catch the local promos and plan to transfer later? Curious if anyone has actually done a y1 to y2 transfer in practice (Porkbun, Namecheap, GoDaddy to anywhere), what the experience was.


Next post: the technical architecture write-up for TubeMine. Server-side broker pattern, monthly per-IP budget on Upstash, and the four things I deliberately left out to ship in 2.5 hours.

tubemine.vercel.app | code (MIT)

Top comments (0)