the cold email problem nobody talks about
when trying to verify tech-blogger emails, we hit a ~50% miss rate. here’s why the target pool for outreach is structurally smaller than it looks, and what it m
recently we’ve been trying to reach out to tech bloggers. writers who cover ai, startups, tools, the whole ecosystem. it’s a classic growth move: find people who write about things like lucy, send a thoughtful note, see if they’re interested.
but something kept happening. our email verification pipeline was failing. a lot. roughly half the time, we couldn’t even find a valid email address to send to.
so we looked closer. and it turns out the problem isn’t our tools. it’s the landscape.
why emails vanish
a few patterns kept showing up in the failures.
first, the substack effect. a huge number of writers now host on substack. and substack doesn’t show plain-text email addresses anywhere. if you want to contact the author, you have to use their built-in contact button. which means no scraper, no outreach tool, no human copying and pasting an address from a page can get it. the email is hidden by design.
second, cloudflare email protection. you’ve seen this: an email address on a page looks like a bunch of hex code wrapped in a span with a cfemail class. it’s obfuscated to stop scrapers from harvesting it. to decode it, you’d have to run javascript on the page, which most verification tools don’t do. so those addresses appear invalid until rendered.
third, the form-wall. the most established people, think sam altman, dan abramov, derek sivers, don’t show an email at all. they’ve been spammed into oblivion. their contact pages are just forms. you fill it out, hope it gets through. no address to copy, no way to verify it exists.
and lastly, the mastodon mirage. sometimes you see something that looks like an email in a sidebar: [email protected]. but that’s not an email address. it’s a mastodon handle. it looks right, but it’s not. our system flagged those as invalid (correctly), but they’re a sign of how contact methods are shifting.
what this means for outreach
the takeaway is simple: the pool of reachable people via cold email is smaller than it seems.
when you scrape linkedin or google for “tech blogger ai”, you might get 10,000 results. but if half of those have no verifiable email address, or no email address at all, your actual ceiling is 5,000. and that’s before you filter for relevance, interest, or timing.
it’s a structural shrink. it’s not that people don’t exist. it’s that the way we’re allowed to contact them has changed. email is becoming a walled garden, a managed channel. and for good reason: spam is unbearable.
so if you’re planning a cold email campaign, don’t overestimate your top-of-funnel. assume a high attrition rate before you even write a word. budget your time and expectations accordingly. maybe invest more in making each email count, since you’ll send fewer. or think about alternative channels, twitter dms, linkedin inmails, comments, even. (though those have their own limits.)
a quiet admission
this isn’t a problem we can fix. lucy can’t decrypt cloudflare’s hex or bypass substack’s contact forms. and we shouldn’t. these protections exist for a reason.
it’s just a reminder that growth is messy. channels change. what worked last year might not work now. and sometimes the data you have isn’t the data you need.
so we adjust. we focus on the ones we can reach. we write better emails. we respect the walls when we find them.
maybe that’s how it should be.
if you’re building something and trying to get the word out, maybe start by checking out who’s already talking about it. like the companions people are building with lucy.
thanks for reading. if this resonated, the product is downstairs.