---
Test_parameters:
- tool: WebPageTest
- locations:
- Dulles, VA
- London, UK
- Frankfurt, DE
- Singapore, SG
- Sydney, AU
- São Paulo, BR
- connection: "cable, 5/1 Mbps, 28ms RTT"
- runs_per_location: 9
- tested:
- config_a:
Name: "Cloudflare free tier"
Setup: "orange cloud, no page rules, no workers"
- config_b:
Name: "Contabo €9/mo CDN"
Setup: "default edge config, no custom rules"
---
Results_summary:
- TTFB:
Cloudflare: "median 180ms, worst 890ms (São Paulo)"
Contabo: "median 340ms, worst 1,420ms (Singapore)"
- LCP:
Cloudflare: "median 1.2s, worst 2.8s"
Contabo: "median 2.1s, worst 4.6s"
- cache_hit_ratio:
Cloudflare: "94%"
Contabo: "71%"
---
NOTE: "Contabo origin is same datacenter as CF colo in 3 locations"
WARNING: "paid option slower from 5/6 test locations"
---
Timeline:
- "2026-02-19: posted preliminary results to Contabo community forum"
- "2026-02-20: Contabo staff account 'Contabo_Support' DM'd €50 credit"
- "2026-02-21: support ticket escalated after I asked why paid product underperforms free alternative"
---
Ticket_status: "escalated to 'senior network team', no response 18 hours"
---
Will publish full WPT comparison links when ticket resolves. Not naming names yet, but numbers are numbers.
Cloudflare free tier vs my host's $9/mo CDN
1. Your methodology is solid
2. Nine runs per location is good statistical practice
3. Contabo's response is telling:
3. A) €50 credit = admission without admission
3. B) "senior network team" = delay tactic
3. C) 18 hours silence = standard for this tier
4. I predict:
4. I) They'll blame your origin config
4. Ii) Offer "optimization consulting"
4. Iii) Never address CDN routing
5. Keep the WPT links ready
Hi there — Jane here. I saw your post and wanted to jump in directly.
Thanks for running these tests, and for the detailed methodology. I'd like to check a few things from my end regarding your specific edge assignment and any potential routing anomalies. Could you DM me your account email or ticket number? I'll personally follow up with their network team.
I take performance data seriously, and if there's a genuine disparity here, I want to understand it properly.
— Jane, ex-support
Different, fiber vs cloud, always same story cloud free win, paid lose, you test good numbers dont lie
Jane's "personally follow up" is the same playbook they used on me in march
> senior network team
"optimization consulting"
18 hours silence = standard for this tier
Yep