The cold DM playbook that still works

Every scalable channel requires signal that doesn't exist yet — authority for SEO, data for ads, an audience for content. Outreach is the only channel that works at true zero, and it converts at rates nothing else touches: 15-30% for well-targeted, genuine messages, versus low single digits for anything cold at scale. The reason is targeting precision — you're choosing exactly who gets the message, one person at a time.

It also does double duty. Every conversation is market research — you learn the language users describe the problem in, which feeds your positioning, your listings, and eventually your content. The founders who do outreach are running a distribution channel and a research program at the same time.

Bad outreach fails on the who, not the what. The message matters less than finding someone who actually has the problem right now — and those people advertise it. They're complaining about the problem on X and Reddit, asking for recommendations in communities, using visible workarounds (a spreadsheet that should be a tool, three products duct-taped together).

Build the list from public pain: 'anyone know a tool that does X' posts, complaints about the competitor everyone hates, workflow complaints in your niche's communities. Twenty people who described your exact problem is a better list than two thousand scraped emails.

The template that works is short, specific, and honest about the ask. One line proving you read them — their actual words about the problem, not 'I love your content.' One line about the thing — what it does, not a pitch deck. One line of ask — not 'hop on a call,' but something they can answer in ten seconds: 'want the link?' or 'is this the kind of thing you'd use?'

The biggest single mistake: pitching to everyone instead of offering to help the few. 'Try my product' is an ask. 'I noticed you were hitting this exact problem — I built something for it, want me to send it?' is an offer to someone who's already in pain.

Twenty DMs a day for ten days is the actual prescription — because day-10 cumulative numbers look nothing like day-3. Most founders send thirty messages, get two replies, and quit inside a week. The channel works at sustained volume and fails at sample size. Two hundred well-targeted messages is a real test; thirty is anxiety dressed up as data.

And measure the right thing: not open rates (invisible on most platforms) but reply rates and conversation rates. A 10% reply rate on targeted outreach is mediocre; 20-30% is the norm for genuinely well-aimed messages. If you're below 10%, the list is wrong before the message is.