The thing I kept from Rachel Konrad’s two Draper U sessions was not a writing trick.

It was something more awkward than that.

At one point she described the sort of CEO she dreads dealing with. Not the nervous one. Not the one who is clumsy on stage. The one heading into an investor meeting, a press interview, or an important partnership conversation and still saying, “I’ll just answer naturally when I’m there.”

People laughed.

I did too. Then it landed a bit differently.

What makes that line funny is not that it is absurd. It is that it is common. A lot of people still think personal brand and cold outreach are communication problems. As if the world would become more receptive the moment the prose got cleaner, the email got smoother, the bio got neater. Once you have done this a few times, the order looks different.

You are not understood first and replied to second.

You are judged first.

Only then do you get the chance to be understood.

Personal brand is not self-expression. It is classification.

Rachel is unusually well placed to talk about this. She teaches at Stanford GSB, started in journalism, and then worked inside companies like Tesla and Impossible Foods where attention is scarce and everyone forms an opinion quickly. People with that background tend to become almost pathologically alert to how others sort you in the first few seconds.

She compressed a fair amount of business writing into two lines: know your audience, know your message.

The problem with advice like that is precisely that it sounds obvious. The more obvious it sounds, the easier it is to assume you already do it.

I used to think about personal brand in a more inward way. Make the background coherent. Explain what you have done. Answer the question of who you are properly. After reading enough founder bios, LinkedIn About sections, fundraising notes, and cold emails, that stopped feeling like the right centre of gravity. The useful part of a personal brand is not completeness. It is whether it lets someone make a correct first judgement quickly.

That is not just a communications point. Social perception research has been circling the same basic structure for years: people tend to evaluate others very quickly along two broad dimensions, warmth and competence. In plainer language, they tend to ask some version of: are your intentions safe, and are you capable enough for me to care?

So a lot of bad introductions do not fail because they lack information. They fail because they are expensive to process.

The CV is long. The About section is full of adjectives. The email is polished. But the reader still cannot tell what kind of problem you are attached to, why you are in their inbox, or what category to file you under. Nothing is obviously wrong. It is just weak tea.

The personal brands that actually work tend to do three things first.

First, they make it easier for people to remember you when a certain class of problem appears.

Second, they let someone describe you in a sentence that is more useful than your title.

Third, they make your direction legible. Not everything you have done, but where you are clearly going.

That is why I no longer think of a LinkedIn About section as the prose version of a CV. The useful version is the one that reveals the logic underneath your choices. Why you keep choosing similar kinds of work. Why the same themes keep recurring. Why the apparently scattered career is not actually scattered.

That is when the thing starts to feel like a point of view rather than a scrapbook.

A cold email is a tiny decision interface

Cold outreach is just the compressed version of the same problem.

Rachel’s second session was nominally about cold emails. What stayed with me was not a subject-line trick. It was her impatience with emails that looked competent on the surface but gave the recipient no good reason to care.

That distinction matters. A lot of people think they are doing outreach when they are really doing distribution. They are spraying messages around and calling the activity effort.

I no longer see cold email as a copywriting problem. It is closer to interface design. The other person needs to answer four questions almost immediately:

  • Who are you?
  • Why me?
  • Why now?
  • If I reply, how annoying will the next step be?

If the answer to any of those remains blurry, the message usually dies there.

This is also why surface-level personalisation rarely impresses me. Putting someone’s name in the subject line or mentioning a recent post is not research. Research is figuring out what they are actually paying attention to right now and then shaping your request around that context.

Rachel was quite specific on this. She talked about looking at what a person is doing across Google News, LinkedIn, X, and podcasts. Not in a creepy way. In a relevance way. People reveal different things in different formats. A press quote is not the same as a thoughtful post. A podcast appearance usually says more about what is top of mind than a bio ever will.

I also liked her line on where the creepy boundary actually is. Public information, voluntarily shared, is fair game. Once you drift into physical-world surveillance, the whole thing curdles. That boundary matters more now than it used to.

Most cold emails fail because they ask for trust too early

This is the part people most often get wrong.

They do not write too little.

They ask for too much.

Please review the deck.
Please give me an hour.
Please introduce me to your partners.
Please advise on the business model.
Please schedule a meeting.
Please tell me what you think.

None of those requests are inherently unreasonable.

They are usually just premature.

The asks I trust more now are almost boringly small.

May I send a one-pager first?
If this overlaps with the thesis you have been posting about, could I ask for 15 minutes?
If this is not your lane, could you point me to the right person?
Would you be open to telling me if I am even framing the problem correctly?

That is not just politeness. It is interface design again. You are lowering the cost of the first yes.

The logic is not far from how early investors look at pitch decks. YC’s seed-deck guidance is useful not because it is sacred, but because it is brutally honest about sequence. Let the reader know what the company does, what problem it solves, why the timing matters, and why this team has a right to exist. Sequoia’s old business-plan guide works for the same reason. It gets the skeleton on the table first.

A cold email should do the same. It is not trying to complete the sale. It is trying to earn the next unit of attention.

AI is most dangerous exactly where the stakes are highest

Rachel’s AI take was one of the more sensible ones I heard that week. She was not anti-tool. She was clear about where the tool should stop.

Her analogy was that large language models are like interns. Fast, willing, often surprisingly useful. But if the writing is carrying something expensive, you do not take the first draft and send it straight to the investor.

That is roughly where I have landed too. AI is genuinely useful for:

  • re-cutting a message for different audiences
  • tightening structure
  • spotting vagueness
  • stripping out clumsy sentences

What it should not decide for you is the positioning layer. If you outsource that too early, you often get something that is perfectly serviceable and almost impossible to remember.

That is not because the model is stupid. It is because the expensive part is not the sentence. It is the judgement. What drawer do you want to be placed in? Which part of you should be legible first? What is the exact next step you are asking for?

If those choices are still fuzzy, the tool will simply make the fuzziness look more polished.

The order that feels least romantic is usually the right one

When I rewrite a LinkedIn About section now, or a cold email that actually matters, I do not start by asking how to make it sound better.

I start here instead:

  • What is this person actually paying attention to right now?
  • What real relationship do I have to that problem?
  • How do I want them to place me before they know me properly?
  • What is the smallest next step I genuinely want?

Once those are clear, the writing usually gets easier.

Because the expensive thing was never the prose.

It was the judgement.

People need to know where to place you before they reply.