Experiment: Running identical sites on .com vs .co for 6 months

So someoneelsesidea finally got me off my seat and running this someoneelsesexperiment. Two identical static sites, same Nameserver, same KnownHost VPS in Amsterdam, same Cloudflare settings disabled[1]. One on projectalpha.com, one on projectalpha.co. Ran from February through July 2026.

Traffic Results:
- Organic Search: .com won handily. 34% more Sessions. Google somehow prefers the .com despite identical Content.
- Direct Traffic: .CO won on mobile. 18% more mobile direct. My theory? Younger users type .co instinctively from all the startup branding.
- Referral: dead even.
- Email Campaign: .co tanked. 23% lower open rate, 41% more spam folder placement[2]. This is the real Story here.

Trust Metrics:
- .com got 12% more Newsletter signups.
- .co had higher bounce rate on desktop.

Footnotes:
[1] yes I disabled Cloudflare. Fight me.
[2] tested with Mailgun and selfhosted Mailcow, same result. Its the TLD reputation.

Dig +short projectalpha.com A
dig +short projectalpha.co A

Two server, same IP, will use same config. The mobile direct result is INTERESTING. I will check my .co property next week, maybe same pattern XD

grep "direct" /var/log/nginx/access.log | awk '{print $1}' | sort | uniq -c

This command, will show you mobile vs desktop? No, need user-agent parse. Berlin1, you use goaccess or custom script?

Hey folks, friendly reminder to keep discussion focused on the experiment itself. @Berlin1 appreciate the detailed writeup, though footnote [1] reads a bit confrontational — edited your tone slightly, not the data.

If anyone wants to replicate, please share your registrar and hosting setup so we can control for variables. The email deliverability angle is genuinely useful for the community.