At IPW 2026, the U.S. Travel Association’s trade show for inbound travel to the United States, every buyer could submit up to 50 appointment requests for a maximum of 42 appointments over three days. The scheduling system then worked through a fixed priority order. Mutual requests between international buyers and exhibitors came first, followed by buyer-driven and then exhibitor-driven requests, and so on down to journalists and committee chairs (IPW 2026 appointments page). Requests were also matched on booth proximity, which keeps buyers from crossing the hall between two back-to-back appointments.
None of that was decided by software. People at U.S. Travel wrote those rules, and the software applied them.
That set of rules has a name in event matchmaking: the meeting matrix. Every pre-scheduled B2B meeting programme runs on one, whether the organiser calls it that or not. In this guide we show you what a meeting matrix contains, how the two grids inside it work, and how to build your first one before booking opens.
Our point is simple. A meeting matrix is a set of decisions you make on paper, not a feature you switch on. Most of what looks like a matchmaking problem on the day is a matrix decision nobody made.
What is a meeting matrix?
Before you build one, it helps to know exactly what the term covers. A meeting matrix is the set of rules that decides which participant groups at a B2B event may meet, in which direction they may request meetings, how many meetings each group can have, and in which time slots and at which tables those meetings take place. It turns registrations into confirmed one-to-one meetings with a time and a place.
In practice it consists of two grids:
- A permission grid: participant groups along both axes, and in each cell the rule for that pair. May buyers request suppliers? May suppliers only accept? May suppliers see each other at all?
- A time grid: time slots along one axis, tables or meeting rooms along the other. Each confirmed meeting occupies exactly one cell, and no person can be in two cells at once.
The term is also used elsewhere. You will find “meeting matrix” templates for prioritising internal meetings, and a room diagramming product for hotel function space carries the same name. This guide is about the planning principle for B2B events: buyers and suppliers, investors and founders, research groups and companies.
In event software, the rules come first and the schedule follows.
The permission grid: who may meet whom
The first grid is where most of your decisions live. Start by listing your participant groups, then fill in each cell with one of three answers: may request, may only view, or hidden.
Here is a simple example for a trade show with a buyer programme:
| Requests from ↓ / to → | Hosted buyers | Exhibitors | Trade visitors | Speakers |
|---|---|---|---|---|
| Hosted buyers | Hidden | May request | Hidden | Hidden |
| Exhibitors | May request | May only view | May request | Hidden |
| Trade visitors | Hidden | May request | May only view | Hidden |
| Speakers | Hidden | Hidden | Hidden | Hidden |
Read it row by row. Hosted buyers may request exhibitors, because that is what their trip is paid for. Exhibitors may request buyers, but they can only view other exhibitors, so the show does not turn into a sales floor between competitors. Trade visitors never see the hosted buyers at all. Speakers stay out of the open search altogether; if a sponsor wants time with one, the programme team arranges it.
The grid does not have to be symmetrical, and usually it should not be. Direction matters. At many buyer programmes, buyers request and exhibitors only accept, because buyers are the scarce side. At an investor conference, founders request and investors decide.
Every empty cell is a decision you postponed. Postponed decisions get made by whoever registers first.
Keep the number of groups small. Four to six groups is enough for most events. Each new group adds a row and a column, so ten groups means a hundred cells to decide and explain. Our older piece on who should meet whom shows how quickly a mixed audience of buyers, developers, recruiters and press needs this kind of separation.
Quotas, caps and priorities
Permissions say who may meet. They do not say how many meetings anyone gets, and that is the question your head of sales will be asked first. Somebody sold exhibitor packages with a promise attached, such as “at least ten pre-scheduled meetings”. The matrix has to keep that promise, or the next sales season starts with refunds.
Three kinds of numbers belong in the matrix:
- Minimums: the number of meetings a group is guaranteed or obliged to hold. Hosted buyer programmes usually set a minimum for buyers in return for their travel.
- Caps: the maximum number of requests or meetings per person. Caps protect popular participants, such as the three largest buyers everyone wants, from a flood of requests.
- Priorities: which requests are scheduled first when two compete for the same slot. Mutual requests, where both sides asked for each other, are the usual first rule.
IPW shows all three in one programme.
Appointment rules at IPW 2026
The minimum of 25 is the deal behind the hosted buyer model: the organiser pays for the room, the buyer fills the diary. The cap of 50 requests is higher than the 42 slots on purpose, because some requests are always declined. For your own event, the consequence is to set the minimum from what your packages promise, and the cap from what your most popular participants can bear. How many meetings a buyer should hold is a question of its own, covered in how many meetings each hosted buyer should get.
Priorities need one more thought. If you rank international buyers above domestic ones, as IPW does, say so in the participant information. Rules that participants can read are rules they accept. And if the two sides of your market are badly out of balance, no priority order will fix it; our guide to the buyer-to-seller ratio shows where that balance usually sits.
The time grid: when and where
Once you know who meets whom and how often, the second grid decides whether it fits. Its arithmetic is simple. Multiply the number of tables by the number of slots, and you have the maximum number of meetings your event can hold.
Take a worked example. You plan 40 tables in a meeting area. Slots last 20 minutes, with 5 minutes to change tables, and you run six hours of meetings a day. That gives 14 slots per table and 560 meetings per day at most. Now your head of sales mentions that 120 exhibitors were each promised eight meetings a day. That is 960 meetings. They do not fit, however clever the software is.
You have three ways out: more tables, more hours or a smaller promise. All three are cheaper before the packages are sold than after.
The time grid also has rules of its own. At Slush 2026, each meeting gets a dedicated table for 25 minutes, followed by a five-minute break, and a slot only shows as busy once a meeting has been accepted, not while a request is pending (Slush 2026 meetings information). Slush also keeps ten tables free for ad-hoc meetings. That small reserve absorbs the meetings people arrange in the corridor.
Researchers have studied this grid as well. A team of computer scientists around Miquel Bofill at the University of Girona calls it the business-to-business meeting scheduling problem: given the pairs that should meet, find times that respect each participant’s availability and the room’s capacity, while minimising idle gaps in participants’ schedules (Bofill et al., IJCAI 2023). For you, the practical lesson is the last part. A diary with three meetings spread over six hours feels empty, even if every meeting is right.
How the matrix turns into schedules
With both grids in place, the matrix still needs a booking mode to produce actual diaries. Organisers mostly choose between three: participants request and accept meetings directly, participants rank whom they want to meet and the organiser generates the schedules, or the organiser approves each request before it is confirmed.
The matrix works the same way in all three. It limits what each person can see and request, and the time grid places every confirmed meeting in a free slot at a free table. The differences between the modes, with real examples and slot lengths, are set out in our guide to B2B meetings at events.
The categories that drive the matrix also have to be collected somewhere. The Virginia Department of General Services runs its 2026 Procurement Forum on commodity codes: small businesses enter NIGP or commodity codes when they register, and the organisers use them to assign the one-on-one schedules (Virginia DGS, Forum 2026 matchmaking guide). Your registration form plays the same role. If a question that drives a matrix rule is optional, that rule fails for everyone who skipped it. Which questions earn their place is covered in registration questions that make matchmaking work.
Build your first meeting matrix in six steps
Everything above fits into a short sequence. Do it in this order, and write each step down before registration opens.
Building a meeting matrix before booking opens
- 1 List your participant groups Four to six groups, each defined by one registration field such as role or package.
- 2 Fill the permission grid For every pair, decide: may request, may only view or hidden. Leave no cell empty.
- 3 Set minimums and caps Translate package promises into minimums, and protect popular participants with request caps.
- 4 Size the time grid Tables times slots per day must cover the sum of all minimums, with a reserve.
- 5 Choose the booking mode Direct requests, ranked preferences with generated schedules, or organiser approval.
- 6 Publish the rules Tell participants who can request whom and how slots are assigned, before booking opens.
Step four is the one most often skipped, and it is the one that decides whether the promises hold. Run the capacity check with your head of sales in the room. Then the number on the package and the number of tables come from the same spreadsheet.
How long this takes in a real project depends mostly on how quickly the rules are agreed, as our realistic matchmaking setup timeline shows.
Common meeting matrix mistakes
Even a written matrix can go wrong. These are the patterns we see most often:
- Everyone may request everyone: it feels open and fair, but the most popular participants drown in requests and everyone else waits. Use direction and caps instead.
- Too many groups: ten groups create a hundred cells, and nobody can explain the result. Merge groups that follow the same rules.
- Rules changed after booking opens: every change affects meetings that are already confirmed. Agree the matrix first and freeze it for the booking window.
- Promises that the time grid cannot hold: the capacity check comes before the price list, not after the first complaint.
- No hidden option: speakers, sponsors’ executives and VIP buyers need a way to stay out of the open search without leaving the event.
Each of these can be fixed. Fixing them before the event costs an afternoon.
Solution: Converve’s matchmaking runs on a rule-based meeting matrix. You define the participant categories and decide for each category pair who can request whom, who can only view whom and who stays hidden. Participants then either send requests and pick time slots, or rank whom they want to meet so that the Meeting Generator builds complete schedules for you to review and publish. Confirmed meetings are scheduled without clashes, including table and room assignment, and because every rule is visible, you can explain any diary to any participant. See how it works on the B2B matchmaking platform page.
Conclusion: decide the matrix before the software applies it
A meeting matrix is not a hidden algorithm. It is two grids of decisions: who may meet whom, in which direction and how often, and when and where those meetings can take place. Write the permission grid, set minimums and caps from what you have promised, and check the time grid against those promises before booking opens. When an exhibitor asks why his diary looks the way it does, you can then point to a rule rather than to the software.
That answer also protects your head of sales. The promise on the package was checked against the tables before it was sold.
If you want to sketch a meeting matrix for your next event, with your participant groups and your capacity, get in touch with Converve.
Frequently Asked Questions
What is a meeting matrix at a B2B event?
A meeting matrix is the set of rules that decides which participant groups may meet, who may request meetings from whom, how many meetings each group can have and in which time slots and at which tables the meetings take place. It turns registrations into confirmed one-to-one meetings, for example between buyers and suppliers at a trade show.
How do you create a meeting matrix?
List four to six participant groups and decide for every pair whether one group may request, only view or not see the other. Then set minimums and caps per group, check that tables times slots cover all promised meetings, choose a booking mode and publish the rules to participants before booking opens.
How many meetings can a B2B event schedule?
Multiply the number of tables by the number of slots. With 40 tables, 20-minute slots plus 5 minutes changeover and six hours of meetings, the maximum is 560 meetings a day (own calculation).
Who decides who meets whom in B2B matchmaking?
The organiser does, through the meeting matrix. At IPW 2026, for example, the U.S. Travel Association set request caps, a minimum of 25 attended appointments for hosted buyers and a priority order that scheduled mutual requests between international buyers and exhibitors first (IPW 2026 appointments page). Software then applies those rules to fill the diaries.
Is a meeting matrix the same as AI matchmaking?
No. A meeting matrix is rule-based: every pairing traces back to a rule the organiser wrote, such as “buyers may request exhibitors”. AI matchmaking is a different approach that ranks suggestions with a learned model. Even where an event uses such a model, the matrix still sets who may meet whom, how often and when.