What thirty minutes of research actually contains.

The first time someone asked me what "thirty minutes of research" really meant, I gave a bad answer. I said something about "going beyond the bio" and watched the prospect's face go flat. She'd been pitched that line by four agencies that month. I had three minutes left in the call and I'd just sounded like all of them.

So let me try to say it properly this time.

Thirty minutes is not skimming more of the same things an SDR would skim.

If it were, it wouldn't be worth doing. An SDR who skims for ten minutes instead of three still ends up at the same opener, because the opener is determined by what the reading surfaces, and skimming surfaces the same thing every time. Thirty minutes is a different sequence. The sequence matters more than the time.

Here's how mine actually goes.

The first four minutes — her posts.

The first four minutes I spend on her posts, not her bio. The bio is where she wrote for strangers. The posts are where she wrote for people she already knows, which is where her real preoccupations leak out. I read three of them, in reverse. One post is an opinion. Two is a position. Three is a preoccupation. I'm looking for the word that shows up in all three. That word is what's actually on her mind this quarter, whatever her title says she should be focused on.

This is the part where I have to be honest with myself. The first time I tried this, I'd find the word in post one and stop. The other two posts felt like padding. They are not padding. The third post is what tells you whether the word from post one was a one-time thing or a real preoccupation. I've written messages off post-one alone and watched them land like nothing, and I know now it's because the word I built the message around was something she'd said once and moved on from.

Minutes four to eleven — the company.

Minutes four to eleven I spend on the company's most recent public thing. The investor letter, the press release, the board change, the rebrand, the hiring page. Whatever's freshest. I'm not reading it to know the company. I'm reading it to find the gap between what the company is saying publicly and what she's preoccupied with personally. Those two things almost always disagree, and the disagreement is the message.

Here's an example of what I mean, even though I won't name anyone. Imagine a head of marketing whose last three posts are all variations on "we're rebuilding brand consistency from the ground up." Then I pull up her company's press releases from the last quarter, and the same product is being described three different ways across them. She's not talking about brand consistency because she's read a book about it. She's talking about it because she's losing the internal argument about it, and she can hear herself losing it every time a new press release goes out. That gap, between her public preoccupation and her company's public output, is the thing nobody else writing to her this week is going to notice. That gap is what gets the reply.

The next seven minutes — a podcast.

The next seven minutes go into hunting for a podcast appearance. Not because podcasts are magic. Because a podcast is the one place senior people stop performing. The first ten minutes of any podcast are the prepared answers — the ones she's given before, the ones she has clean. I skip ahead. I listen at one-and-a-half speed to the middle of the episode, where the host has stopped doing the warm-up and started asking what she's actually working on. I'm listening for the moment her cadence breaks. The pause before a sentence. The half-laugh. The "honestly, I don't know yet." That's the moment she's thinking out loud, and it's the one moment in her public footprint where she's not running her words past the version of herself she's calibrated to present.

If there's no podcast, I'll take a long-form interview, a conference Q&A, a panel where she had to speak for more than ninety seconds at a stretch. The format doesn't matter. What I'm hunting for is the place she ran out of her prepared script.

Minutes seventeen to twenty-two — discarding.

Then, between minute seventeen and minute twenty-two, I do the part nobody talks about. I throw away the first three angles I came up with.

This is the part I had to learn the hard way. By minute seventeen I already have three perfectly reasonable angles for the message. They are all wrong. They are wrong because they are exactly the angles a different person — a smarter SDR, an AI tool, another agency doing two minutes of research — would also surface. The whole point of paying for thirty minutes is to get past the obvious, and obviousness has a specific texture. It shows up as the angle that you arrived at in the first ten minutes of reading. If I send that angle, I've spent thirty minutes to write a three-minute message, and the prospect will read it as a three-minute message because that's what it is.

So I sit with the discomfort and throw them out. The right angle is almost always the fourth or fifth one, and it almost always requires holding two of the earlier facts next to each other. The podcast moment plus the hiring page. The repeated word plus the press release. The contradiction between what she's saying publicly and what her company is saying publicly. The right angle is the one you can only see if you've done all four of the earlier steps and then refused to use any of them in isolation.

This is the part of the work that nobody can see in the output. The throwaways. The minutes I spent on angles that don't appear in the final message. If you only saw the message, you'd think I read for ten minutes. You can't see the twenty minutes of reading I did for the angle I didn't use.

The draft, then the cut.

The draft happens in the next six minutes. One opener. Around 140 words on the first pass. I write the whole thing around sentence five.

Most cold messages fail at sentence five, not sentence one. The opener can be specific and personalized and the prospect will still archive the message if sentence five is templated. She doesn't always know she's doing this. She just feels the message turn generic and she stops reading. So I write sentence five first. It's the one that has to carry the claim that I actually read enough to notice the thing nobody else noticed. The opener gets one specific reference. Sentences two through four set up the reason that reference matters. Sentence five lands the angle from minute twenty-one. Everything after sentence five is short and clean and easy to reply to.

The last two minutes I cut. 140 words to around 70. Every adjective goes. Every "I wanted to reach out." Every "I noticed you're focused on." Every "I'd love to learn more about." Anything that sounds like an SDR being trained out of a playbook. Anything an AI would put in because it's been trained on a million versions of the same warm-up sentence.

What's left is short enough to read in one breath and long enough to do one job — show her, in a way she can verify in three seconds, that someone read enough to notice the thing.

That's thirty minutes.

The point isn't the time.

I want to say one more thing about it, because the time isn't really the point. The point is the refusal. The refusal to send the obvious angle. The refusal to use the bio. The refusal to write sentence five the way it would write itself. Thirty minutes is what it takes to keep refusing all the way to the end of the message. Ten minutes isn't enough to refuse. You arrive at the obvious angle, you don't have the time to throw it out, and you send it. Three minutes is just sending the obvious angle on purpose.

Nobody in a healthy SDR org has thirty minutes per prospect. The math doesn't work. That's fine — it shouldn't. I write fewer than ten messages a day. I charge what I charge. The thirty minutes is the floor of the work, not a feature I added to justify the price. If I cut the time, the work becomes the same work everyone else is doing, and the prospect can already tell when she's reading the work everyone else is doing.

The math is the moat. Once it stops being the math, there's no moat left.