This teardown is for teams on Webflow weighing seats and workspace plans against an owned Astro site. By the end you’ll have the ~$113/mo two-person sketch, the three-year compounding frame, and a clear stay-vs-leave test, using invoice math, not framework fandom.
Use the free own vs rent comparator with your actual monthly Webflow + seat line items. For the broader marketing-stack invoice that often sits next to Webflow, see Webflow + Zapier + HubSpot: the real monthly cost.
01 · What each platform sells you
Different products, different landlords
Webflow sells a hosted design-and-CMS platform. You rent the canvas, the hosting, and the collaboration model. The site plan is the memorable sticker. The workspace and seats are how the bill grows.
Astro on Cloudflare sells a static-first site architecture you deploy as software you own. Hosting for a typical marketing site is near zero on Cloudflare Pages. There is no per-editor seat meter on the host. Collaboration happens in git, a CMS you choose, or both, not in a platform seat matrix.
Neither is “more modern” in a way that matters on the P&L. One meters collaboration. The other front-loads a build and then flattens.
02 · Seat cost
The $25 site is not the team price
Webflow’s Premium site plan at roughly $25/month (annual) is the number that sticks. Solo operators with one site often land close to that. The moment marketing needs a second person in the canvas, the model shifts:
- Growth workspace: about $49/month
- Each additional full (“designer”) seat: about $39/month
- Two-person editing sketch: closer to ~$113/month before add-ons
That is still reasonable relative to an agency retainer. It is not the sticker. It also does not buy you the repository.
| Sticker. Premium site planthe number people remember | ~$25 |
| Premium site plan | ~$25 |
| Growth workspaceteam collaboration | ~$49 |
| 1 additional full seat | ~$39 |
| Team monthly sketch | ~$113 |
Astro on Cloudflare has no equivalent seat line. Editors either work in a CMS you pay for once (often cheaper than stacking Webflow seats) or ship through pull requests. The host does not charge you every time someone needs write access to the marketing site.
03 · Three-year TCO
Compound rent vs front-loaded ownership
Multiply the team Webflow sketch by thirty-six months and you already have a real number, before localization, enterprise plans, agency markups, or a mid-cycle redesign. Stack Zapier and HubSpot beside it and the marketing stack becomes the ~$1,500/mo loaded pattern many teams already live in.
| Webflow team seats × 36~$113/mo sketch | ~$4,100 |
| Astro hosting × 36Cloudflare · typical marketing traffic | ~$0–$360 |
| Owned Astro build (once)site, redirects, CMS plan, handoff | $18,000–$40,000 |
| Decision frameseats compound; build flattens | Compare on own-vs-rent |
The owned path looks expensive in month one. On a three-year timeline, seat rent keeps compounding while the owned site’s hosting line stays flat. Plug your retainer or Webflow line into the own vs rent comparator and find the crossover month on your numbers, not a blog’s.
For agency retainers that bundle Webflow babysitting, the three-year math is worse: see agency retainer vs owned site TCO.
04 · Ownership
What you hold at month 36
On Webflow: you hold a project in someone else’s product. Export is a partial escape hatch. CMS collections, form backends, memberships, and integration glue do not leave as a clean system of record.
On Astro: you hold a git repository, a deploy account (Cloudflare), and whatever CMS or content collections you chose. Agents and humans can change the same codebase. The public site is an artifact of software you control, not a tenant in a design SaaS.
Ownership is the quiet line item. Teams notice it when they want to leave an agency, change vendors, or automate the site with coding agents. Framework preference is noise next to that.
05 · Decision frame
When Webflow still fits, and when it does not
| Constraint | Lean Webflow | Lean Astro + Cloudflare |
|---|---|---|
| Daily visual canvas for marketers | Strong | Only with CMS/editor plan |
| Seat count climbing past two full editors | Bill compounds | No host seat tax |
| Need agents editing real UI code weekly | Awkward | Natural |
| Care about three-year TCO + exit | Weak | Strong |
| Already stacked Zapier + HubSpot | Site is one rent among many | Cut the site rent; fix middleware next |
Stay on Webflow when the canvas is the product and the seat bill is still small. Move when you are paying for collaboration meters, agency CMS babysitting, or you need a repo AI agents can maintain without fighting a proprietary canvas.
This is not anti-Webflow craft. Webflow is good software. It is a landlord model. Astro on Cloudflare is an ownership model. Pick the one that matches how you want month 36 to look.
Conclusion
Price the seats, then price the exit
Webflow vs Astro for a business site is not a DX holy war. It is whether collaboration should be a monthly seat product and whether you want code ownership when the vendor, the agency, or the plan changes.
Run the seat line and the build estimate through the own vs rent comparator again if your invoice changed. Then choose the path that still makes sense when someone asks who holds the keys.