Launched Signal Brief today. First 6 hours, honest numbers.
Sent Issue #1 this morning to 27 subscribers — 25 friends and family, 2 genuine.
6 hours in:
Open rate: 20%
Click-to-open: 60%
Unsubscribes: 0
Problem: landing in spam
The spam issue was expected — new domain, no sender reputation yet. Asking subscribers to mark as Not Spam and hoping it improves over the next few issues.
The 60% click-to-open surprised me. Whoever opened it, clicked. That's the one thing I'll take from today.
Building in public. Week 1 of many. If you've dealt with the spam problem before — what actually helped?
Trending on Indie Hackers
Quick update — open rate is now at 20% with 60% click-to-open. The spam issue is slowly resolving as subscribers mark it as Not Spam.
Biggest lesson from hour 1: sender reputation is built over time, not overnight. Issue #2 is already being planned for June 15.
Good update.
The spam recovery is one part, but I’d be careful with the bigger lesson before Issue #2.
If the next send improves, it will be easy to assume the content is working when part of the lift may just be deliverability normalizing.
The useful thing is deciding what Issue #2 should actually prove, not just whether opens go up.
If helpful, send me your email and I’ll put the tighter read together properly before June 15.
Really sharp point — the deliverability normalization vs content performance distinction is something I need to be more disciplined about tracking separately.
For Issue #2 I'm thinking the hypothesis should be around whether the curation angle (signal vs noise) actually drives replies and forwards — not just opens.
Appreciate the offer — I'll reach out closer to June 15 if I need a second pair of eyes.
That’s a much cleaner hypothesis than treating opens as the whole story.
I’d just be careful not to accidentally validate “signal vs noise” through a metric that can still be distorted by deliverability recovery.
The useful part before June 15 is deciding what would count as a real signal vs a misleading one before the send goes out.
No need to solve it now, but worth tightening before Issue #2.
Fair — I need a cleaner signal than opens or even
forwards if deliverability is still normalizing.
Thinking reply rate might be the most honest metric
for Issue #2 — harder to fake and not affected by
deliverability. Does that hold up?
I wouldn't answer that metric question in isolation.
At this point the useful work is deciding what would constitute a real signal, what would constitute a false signal, and what conclusion each outcome should lead to before Issue #2 goes out.
That's not something I tend to unpack casually in threads because it's easy to end up with fragments of advice instead of a usable decision.
If you want the full version, happy to put it together properly.