Eventira Connect: matchmaking Pilot

The problem it’s solving

Someone at a two-day conference has time for maybe a dozen useful conversations. What most event platforms hand them is a roster: an alphabetical, searchable list of everyone present. Finding the right dozen in that list is manual, luck-driven work done in corridors and coffee queues, and most people never do it. They talk to whoever they sat next to.

That costs three different people something:

• Attendees go to the sessions but don’t meet the people, which is usually why they came.
• You can report attendance, session popularity and message delivery, but nothing about whether the event did the thing it exists to do.
• Exhibitors and sponsors buy a stand and hope. Nothing puts a qualified person in front of them.

What an attendee experiences

1. They opt in

   Nothing about a person is read until they choose to take part, and they can leave at any time. A completeness meter shows how much of the profile they’ve filled in, because a thin profile produces thin suggestions and it’s better they see that than wonder.

2. They say what they want and what they have

   What they’re looking for, what they can offer, why they came, their sector, seniority and company size. The distinction between seeking and offering is what makes the matching directional: your list is built from people who have what you want and want what you have, not simply people who resemble you.

3. They get a ranked list, with reasons

   A For you list with a plain sentence under each name, naming the actual shared terms: “They offer Payments and Compliance”, “You are both in Fintech”. They can save a suggestion for later or dismiss it.

4. They send a connection request

   With a short intro, capped at a couple of sentences. If the other person accepts, the pair get a private chat room.

5. They book a meeting

   Connected pairs propose a meeting in a real room, from a grid of slots you publish. They’re only offered slots that are genuinely free, checked against the other person’s agenda, so an offered slot is an honest offer rather than a guess.

Meeting suggestions: the stronger signal

A ranked list is a browsing surface. It’s one-directional: your list saying “meet Sam” tells you nothing about whether you appear anywhere on Sam’s.

Meeting suggestions only exist where both people rank each other. That’s a different message, and a much easier one to act on: this person wants to meet you too. They’re spread under a per-person budget, so the event ends up with the largest possible number of real meetings rather than a handful of people fielding all of them.

What you set up

• The questions. Which facets your event asks and what the allowed answers are. Two preset vocabularies are available if you’d rather not start from a blank page.
• The weighting. A conference or trade show preset. They differ in ways that matter. At a trade show, what someone offers is weighted above what they’re seeking, and sector counts for more. Seniority counts for much less, because a trade show floor is about fit, not peer level.
• The room grid. The meeting slots attendees can book, with overlapping slots in the same room rejected.
• Registration questions. The matchmaking questions can be added to your registration form in one step, so intent is captured when people sign up rather than chased afterwards.
• Partner groups with outreach quotas, and a consented lead export for exhibitors.

You then trigger a run, and a dashboard tells you whether it’s working: opt-in rate, how complete profiles are, and how many suggestions are turning into real meetings. A cohort report shows groups of people who stated something and got nothing back, which is the most actionable screen in the feature, it tells you which question to add or which vocabulary term is missing.

What no one can do, including you

These are deliberate design decisions rather than gaps, and they’re worth knowing because attendees will ask.

• Nobody sees a score. Not the attendee, not you, not a partner. Scores exist only to order a list. They also aren’t comparable between two people’s lists, which is precisely why they’re never shown.
• A suggestion that can’t be explained isn’t shown. If the engine can’t write a sentence naming the shared terms, the suggestion is discarded rather than displayed without a reason.
• Refusals all look the same. Blocked, unavailable, and gone-private return an identical message. If you could tell them apart, blocking someone would quietly notify them that they’d been blocked.
• You can’t see an individual’s suggestions. Organiser reporting is aggregate only, and any group of fewer than five people is withheld rather than shown.
• You can’t book on someone’s behalf. A published meeting plan is advice. It recommends a slot; the two attendees still have to agree.

Fairness is built in, not left to chance

Left alone, a few attractive profiles absorb an entire event’s outreach. The people with thin profiles need the help most and get the least of it. Connect caps how many lists any one person can appear on, with a floor so that a thin profile still gets a usable list of its own.

There’s a matching piece of care in the scoring: whether a question counts is decided by your answers, never the other person’s. Leave a question blank and it’s dropped entirely and its weight redistributed, so you aren’t punished for your own thin profile, and nobody is rewarded for saying nothing.