The usual way to judge event networking is a survey question: "How satisfied were you with the networking opportunities?", scored from one to five. It is easy to ask and hard to act on. A 3.6 doesn't tell you whether the problem was the matching or the meeting area, the agenda or the fact that half the room never joined.
If your attendees use a networking tool, you already have better evidence. Every request, acceptance, cancellation and missed meeting is a record of what people did, not what they remembered feeling a week later. This guide covers which numbers to track, how to read them together, and what each one tells you to change.
Think in a funnel
Networking is a sequence of steps, and people drop out at each one:
- They have a ticket.
- They join networking and become visible to others.
- They send or receive a meeting request.
- A meeting is accepted.
- The meeting takes place.
- They stay in touch afterwards.
Each step has a number, and each number points to a different fix. A single satisfaction score can't tell you where people dropped out. A funnel can.
The numbers worth tracking
Opt-in rate
People who joined networking, divided by paid tickets. This is your adoption number, and it sets the ceiling for everything below it.
A low opt-in rate is rarely a matching problem. It usually comes down to how people were invited and how easy it was to get in. Our guide to getting attendees to use your event app covers that side.
Share of participants with at least one accepted meeting
Of the people who joined networking, how many had at least one accepted meeting? This is the closest you can get to "did networking work for this person" without asking them.
It matters more than the total number of meetings. A few very active people can produce an impressive total while most of the room does nothing.
Acceptance rate
Accepted requests divided by requests sent. A low rate can mean weak recommendations or requests sent without a note. Sometimes it just means people stopped checking the app after the first day.
Meetings per participant
Accepted meetings multiplied by two, divided by participants. Read it next to the previous metric. An average can hide a small group doing most of the meetings, so calculate it twice: once for everyone who joined, and once for only those who had at least one meeting. Then look at who the most active people were.
Before versus during
The share of meetings booked before the event started. A healthy pre-event share means people planned their time. A share close to zero means your pre-event window didn't work, or never existed.
No-show rate
Accepted meetings where someone didn't turn up, divided by accepted meetings. This is mostly an operations number. It tells you about meeting hours and signage more than about the quality of matches.
Connections and follow-through
Connection requests sent compared with accepted, and whether people kept talking afterwards. It is the slowest signal to arrive, but the closest to the value attendees actually took home.
A worked example
Say a two-day conference ends with these numbers:
- 500 paid tickets
- 310 attendees opted in to networking
- 400 meeting requests sent, 236 accepted
- 170 attendees with at least one accepted meeting
- 60 of the accepted meetings booked before the event, 176 during it
- 20 no-shows
Work them through:
- Opt-in rate: 310 ÷ 500 = 62%
- Share with a meeting: 170 ÷ 310 = 55%
- Acceptance rate: 236 ÷ 400 = 59%
- Meetings per participant: 236 × 2 ÷ 310 = about 1.5
- Meetings per participant who met anyone: 472 ÷ 170 = about 2.8
- Booked before the event: 60 ÷ 236 = 25%
- No-show rate: 20 ÷ 236 = 8.5%
How to read it. The weak point isn't adoption. It is the gap between joining and meeting: 310 minus 170, or 140 people, made themselves visible and left without a single accepted meeting. Those who did meet someone averaged almost three meetings each, so the meetings themselves worked for the people who got that far.
With only a quarter of meetings booked before the event, the likely fixes are a longer and better-promoted pre-event window, and more help on site for people who won't start conversations, such as hosts and roundtables, plus an announcement before each meeting block pointing people to their recommendations. Twenty missed meetings is small enough to look at one by one. Check which slots they fell in and what was on stage at the time.
Reading the numbers together
No single metric tells you what to do. Patterns across two or three of them usually do.
| What you see | Likely cause | What to change |
|---|---|---|
| Low opt-in rate | Invites went out late, or access was confusing | Invite earlier, simplify access, say what people get |
| Good opt-in, few people with meetings | Thin recommendations, or people waiting to be asked | Tighter interest list, goals per interest, hosts and roundtables |
| Low acceptance rate | Weak matches, or requests sent without a note | Better onboarding questions, encourage a line of context |
| Almost nothing booked before | Matching opened late or was never promoted | Open matching about ten days out, announce requests a week before |
| No-shows bunched in certain slots | A clash with a popular session | Move meeting hours around the agenda |
| Many "new time proposed" replies | Meeting hours too narrow | Add a meeting block or extend the hours |
Most of the operational rows come down to meeting hours and the meeting area, which you control directly.
Other signals worth a look
- Top interests and the goal breakdown. If nearly everyone chose Networking as their goal for every interest, the goals weren't doing any work. Tighten which goals each interest allows next time.
- Meetings by status. Accepted, declined, cancelled, pending, completed, no-show and new time proposed. A large pending pile at the end of the event means requests nobody answered.
- Most active people. Useful for spotting a sponsor who over-sent, and for finding regulars worth inviting as hosts next year.
- Sponsor meetings. If sponsors paid for introductions, count them. It is the first number they will ask for.
Conecta's organizer analytics report each of these alongside the funnel numbers above, so you are not rebuilding them in a spreadsheet the morning after.
Set targets before the event, not after
Numbers are easier to read when you wrote down what you expected beforehand. Two sources make targets realistic.
The first is capacity. If you have 24 tables, 20-minute slots and four meeting hours, you have 288 meeting slots for the day. With 300 participants, that is a ceiling of just under two meetings each, so a target of three per participant can't be met however good the matching is. Do the capacity sums first, then set the target. Our guide to running one-to-one meetings at events walks through them.
The second is last year. Compare like with like: the same metrics, calculated the same way, for a similar audience. If the ticket mix changed a lot, say twice as many students, note it next to the numbers so nobody reads a drop in acceptance as a failure of the matching.
A target for each of the first four metrics in the report below is plenty. More than that and nobody remembers them on the day.
What doesn't measure networking
Some numbers look like evidence and aren't:
- App downloads. They count intent to look, not meetings.
- Total messages sent. One enthusiastic sponsor can send hundreds.
- Profile views. Browsing is not meeting.
- A single satisfaction score. It tells you the mood, not the cause.
These are not useless. They just answer a different question from "did the right people meet".
A one-page report
After each event, put these on one page and keep the same format from year to year:
- Opt-in rate
- Share of participants with at least one accepted meeting
- Acceptance rate
- Meetings per participant, for everyone and for those who met someone
- Share booked before the event
- No-show rate, and the slots where no-shows bunched
- Connections requested compared with accepted
- One sentence on what you will change, tied to the weakest number
The last line is the one that matters. A report that doesn't change next year's plan is a scrapbook. If you want a sense of what the organizer side looks like before your next event, the page for organizers walks through it.
The takeaway
Measure networking as a funnel, from ticket to accepted meeting to follow-up, and find the step where most people dropped out. Read the numbers in pairs, because an average or a total on its own will flatter you. Keep surveys for what they are good at, which is opinions, and use behaviour for everything else. Then write down the one thing you will change, and check the same number next year.