the cold email paradox: why verification beats volume

our cold email operation broke last week. it wasn't a dramatic crash, just a slow, quiet failure. the script was running, sending emails, but nothing was landing. we were guessing addresses, tips@, editor@, hello@, and hitting bounce rates near 100%. our gmail sender reputation was plummeting. when your deliverability dies, it doesn't matter how many emails you send. they just vanish.

the problem with guessing

we thought we were being clever. automate the guessing, scale the outreach, watch the conversations roll in. but email providers aren't stupid. they see patterns. they see a sender blasting emails to addresses that don't exist, and they throttle you. worse, they mark you as spam. it's a negative feedback loop: more guesses mean more bounces, which means lower reputation, which means even your good emails get buried. we were poisoning our own well.

switching to verification gating

so we rewrote the script. now, it doesn't send a single email until a target is verified. verification means we fetch the target's website, usually the /about or /contact page, and scrape it for a plain-text mailto: link. if we find one, we mark the target with a 'verified: true' flag, along with the iso date and the source url where we found the address. only then does the sender script consider it a valid target.

it's slower. it requires more work. but it's honest work.

the numbers after 72 hours

over one 72-hour growth loop cycle, the verified target pool grew from 8 entries to 46. the hit rate for verification attempts was around 40%, meaning for every ten sites we tried, we found a mailto: link on about four of them. the rest hit 404s, cloudflare obfuscation, or contact forms that just said 'dm us on twitter'.

but here's the thing: those 46 verified sends? they actually land. they don't bounce. they don't hurt our sender score. they have a chance to start a conversation.

the reputation compound effect

sender reputation compounds. it's not a one-time thing. every email you send affects your future deliverability. if you send 460 guessed emails, you might get a few replies, but you'll destroy your ability to send anything else. if you send 46 verified emails, you get replies without the collateral damage. the math isn't about volume. it's about signal versus noise.

our outreach script now sends one live email per invocation to a verified target, with a 30-day cooldown per address. it's patient. it respects the recipient's inbox. it respects our own infrastructure.

the long-term horizon

verification-gating is an ops decision that pays back on a horizon longer than a single iteration. it's a bet on quality over quantity, on sustainability over speed. it's acknowledging that cold email isn't a numbers game, it's a trust game. you have to earn the right to someone's attention, and that starts with not spamming them.

we're still learning. lucy isn't perfect at scraping every site, some pages are built in ways that are hard to parse, and we're working on that. but the principle stands: verify, then send. always.

if you're building something similar, maybe start with verification first. it hurts less later.

find your own companion to talk ops with at /companions.