For Conference Organizers: You Built the Room and Left the Introductions to Chance

You will spend nine months on this event. You'll negotiate the venue, chase the keynote, argue about the track structure, taste four caterers, approve a badge design, and lose a week of your life to the app vendor.

And then you'll ask attendees why they came, and they'll say: the people.

Not the talks. Everyone says the talks are why they came and the people are why they'll come back, and everyone knows the second half is the true half. The talks are on YouTube in three weeks. The people are the product.

Here is the thing worth staring at. Of all the variables you control with obsessive precision, the one that determines whether the event worked — did the right people find each other — is the one you've left almost entirely to chance. You built a room. You put the right seven hundred people in it. And then you stepped back and hoped that proximity would do the rest.

Proximity does not do the rest. Proximity is the easy part.

The hallway track is an admission of failure

We speak fondly about the hallway track, and the fondness is a tell. What we're saying is: the most valuable part of our event is the part we didn't design.

Fine — except the hallway track has a hit rate you'd never accept anywhere else in your operation. Seven hundred attendees. Each one talks to, generously, fifteen people over two days. Of those fifteen, maybe two are meaningful. The other thirteen are polite exchanges of job titles that evaporate before the flight home.

Meanwhile, in that same room, there were probably forty people who each attendee should have met. They stood eight feet apart in the coffee line. Neither had any way to know. They both went home and told you the event was great, because it was — they got their two. They have no idea about the other forty. Neither do you.

You are running a matching market with no matching mechanism. The matches that happen are the ones that survived a random walk.

The app doesn't fix it, and you know it

Every organizer has bought the app. Every app has the attendee directory. Nobody uses it.

The reason isn't UX. It's that the directory contains names, titles, and companies — which is exactly the information that doesn't help. Knowing that someone is "VP Engineering at a Series C infrastructure company" tells you nothing about whether you should spend twenty minutes with them. You need to know what they're working on, what they're stuck on, what they're looking for, and whether you have any of it. None of that is in the directory, because the directory was populated from a registration form, and nobody has ever said anything true on a registration form.

So the app has a matchmaking feature bolted on, and it matches people by tag overlap — you both selected "AI" — which produces suggestions so useless that everyone stops opening it by lunch on day one.

The failure isn't the software. It's the input. You collected the wrong information, at the wrong moment, through the wrong interface, and no amount of algorithm downstream fixes a registration form.

The registration form is the bug

Think about when you ask attendees who they are. Six weeks out, in a checkout flow, next to a field about dietary restrictions. You give them a dropdown for their role and a text box for their company. They fill it in ninety seconds with the least interesting true facts about themselves and click pay.

Now think about what you actually needed to know. What are you working on right now? What decision are you trying to make this quarter? Who would you drop everything to talk to? What would you happily explain to someone for an hour? What are you emphatically not looking for?

Nobody answers those in a checkout flow. But everyone answers them in a conversation — easily, at length, with genuine specificity, if something asks and keeps asking. That's the difference between a form and an interlocutor. A form extracts the shortest legal answer. A conversation extracts the truth.

The information you need to make your event work exists in every attendee's head. You've never had an instrument capable of getting it out.

What the room looks like with context in it

Give every attendee a page that can hold a conversation — before the event, not at it — and the physics change.

Each one has, at some point in the weeks prior, actually talked about what they're doing and what they want. Not filled a form. Talked. That context lives on their page, and their page can be asked questions by anyone, including other attendees, including your systems.

Now the coffee line looks different. Two people eight feet apart, each of whom can find out in twenty seconds whether the other is worth the conversation — not from a job title, but from what each of them actually said they were trying to do.

And you, the organizer, are no longer guessing. You know the actual distribution of intent in your room. Not the tags people picked from your dropdown — the real thing, in their words. You can see that thirty-one people came looking for the same thing and you scheduled nothing for them. You can see the two clusters that never overlap and should. You can put the right people at the right dinner table because you know something real about all of them.

That's programming built on evidence instead of instinct.

The afterlife problem

And then there's the part that actually determines whether people come back.

Two weeks after your event, an attendee is trying to remember the name of the person they met on day two — the one with the specific insight about their exact problem. They have a first name. They have a vague memory of the company. They have nothing else, because they exchanged a business card that's in a jacket pocket, or a LinkedIn connection that's now indistinguishable from four hundred others.

The connection your event created just died. Two weeks. It was the whole value of the event and it lasted fourteen days.

This is not an attendee failing. It's an infrastructure failing. Human memory was never going to hold seven hundred names and the context attached to them, and every tool we've offered — cards, LinkedIn, the app's contact export — moves identity without moving context. Identity without context is a name you can't place.

A rolodex that fills itself as the conversations happen, holding what was actually said, survives the flight home. That's what makes the connection real rather than commemorative.

What you're actually selling

You're not selling talks. You're selling the probability that someone meets the person who changes their year.

Right now you're selling that probability without any ability to influence it. You've optimized every input except the output. The catering is dialed. The introductions are a coin flip.

The room is the easy part. You've already built it. The hard part — carrying enough context that the right people recognize each other before they walk past — is the part nobody has built, and it's the part your attendees would actually pay for.

Parlei is pre-launch. Early access is open to organizers who'd rather engineer the hallway than romanticize it.