In-Person Is the Moat

Here’s a number I can defend: every dinner I’ve ever hosted produces 2 to 3 real opportunities, and those close at roughly 33%.

Here’s the asterisk that makes the number almost useless to a CFO: it takes 9 to 12 months to see any of it.

That lag is the entire story. It’s why most people quit in-person before it pays, why the ones who don’t build something competitors can’t buy their way around, and why I keep getting on trains and planes to stand in rooms with people who aren’t going to buy anything from me, maybe ever.

The receipt no dashboard will print

Let me walk you through an actual chain, backwards.

Last week a founder reached out. We talked the same day. There’s a six-figure deal on the table.

That conversation happened because of an intro call in the first week of April. That intro came from an advisor at a VC firm, who knew me because I’d worked with one of their portfolio companies the year before.

That portfolio company came to us because I’d spent time with their founder at event after event after event. Not one meeting. Not a pitch. A dozen rooms over a couple of years until working together was the obvious next thing rather than a proposal I had to win.

And I met them because in 2023 I took the chance of partnering with Sales Assembly and got on a plane to Chicago.

That action led to a $250,000 contract plus several relationships, which have helped propel my personal and corporate brands forward, and now another six-figure deal on the table.

Now go ahead and model that. Put it in a spreadsheet. Build the attribution report.

You can’t. There is no field in Salesforce for “I sat next to this guy at a dinner in Chicago in 2023 and we talked about having kids.” The system of record captures the last touch, maybe the first, and none of the human infrastructure underneath. I know that chain exists because I lived it, not because a tool told me.

That’s the uncomfortable truth about in-person: the ROI is real and the reporting is a lie. If you need the dashboard to justify the spend, you will kill the program in month four, right before it starts working.

The lag is a feature

Every operator I know is under pressure to compress cycles. Faster ramp, faster time-to-value, faster payback. So a channel with a 9-to-12 month lag looks like a misstep.

Cut it, and here’s what actually happens: nothing, for two quarters. Then a slow, unattributable decline in the inbound you can’t explain, the referrals that stop showing up, the “hey, a friend mentioned you” emails that thin out.

The lag is what makes it defensible. Anyone can outspend you on ads tomorrow. Nobody can retroactively have been in the room with you for three years.

I’ve been doing this for almost six years. Polaris is thirteen months old and did seven figures in that time, all during a year when we had our fourth kid and I had to be mostly out of pocket for a month when my mom passed. We did not have a strong first half. This quarter our win rate is 63%.

That’s not because we got better at selling in July. It’s because the work that was done in the fall of last year, and the year before that, and the year before that, finally showed up on the calendar at the same time. The pipeline was already built. I just wasn’t standing next to it yet.

The bond is never the business

The thing that surprises people: almost none of these relationships were built on GTM talk.

There’s an operator I’ve known for years who I bond with over raising kids and arguing about politics, and we agree on almost nothing politically. Never once talked shop in a way that mattered. She’s in my corner anyway.

When we launched the podcast, every single guest but one was somebody I texted. Not booked. Not pitched through a producer. Texted, the way you text a friend. Because we’d been on the road together, and “I’ll see you in two weeks” is the phrase that turns a contact into a person.

And the episodes that perform best aren’t the sales ones. They’re the ones where we talk about being parents, about life lessons, about how you show up with the same intensity and intentionality as a spouse and a parent that you bring to being an executive. The AI episode did fine. The one about hard lessons learned went viral.

People don’t remember your positioning. They remember whether you were a person.

What persists

Your title is rented. Your equity is contingent. Your org chart will be redrawn by somebody who has never met you. What you actually own, the only thing that ports cleanly across every acquisition, pivot, layoff, and bad quarter: it’s the set of people who would pick up the phone for you.

I can send four names to a friend looking for an operating partner role, with a “tell them I sent you,” and those four calls will get taken. That’s not a favor I’m doing him. That’s an asset I spent six years building, and I get to spend it without spending it down.

So how do you choose?

I over-index on in-person, and I’ll own that. I also have four kids, six and under, and I get them from 8:00 to 9:00 in the morning and 5:30 to 8:00 at night. The cost is real. Anyone selling you a version of this where the tradeoff doesn’t exist is selling something.

Here’s the filter I actually use:

Rooms with repeats beat rooms with reach. A hundred strangers once is worth less than fifteen people I’ll see again in six weeks. Compounding requires the same faces.

Go where you can give something. I make introductions, I host, I connect people to jobs. Not always strategically and the second-order effects are enormous. The person you help in year one is the person who vouches for you in year three.

Partner instead of paying. We run hackathons around the country with VCs and tech partners — 645 in New York, Atlanta Tech Village, Drive Capital & Tech Nexus in Chicago. Shared cost, shared audience, better room than any of us convenes alone.

Judge on nine months, not nine days. If you’re evaluating an event by the meetings you booked at it, you’ve already lost. Evaluate the cohort you’ve been building with over three years.

The actual moat

Everything I do is getting faster and cheaper to replicate. AI writes the sequences, builds the models, drafts the deck, cleans the data. I’m building products in this space myself. I expect just about everyone is.

But nobody can automate having been there.

Nobody can synthesize the fact that we commiserated in a hotel bar about a rough quarter. Nobody can generate the trust that comes from four years of small, unremarkable, in-person consistency. When every output looks the same, the differentiator collapses back down to: do I know this person, and do I trust them?

That’s the moat. It’s slow, it’s expensive, it doesn’t report well, and it takes years before it looks like anything at all.

Then one day someone reaches out cold, and you trace it back four hops to a plane ticket you bought in 2022.

Cliff Simon is the founder of Polaris Ops, an AI-first RevOps consultancy building the systems, data foundations, and go-to-market infrastructure that high-growth companies actually run on.