The simplest matchmaking idea is to pair the people with the most in common. Same sector, same topics. It is easy to build and easy to explain, and at many events it produces the wrong meetings.
Two founders who both picked "Climate tech" and "Fundraising" have plenty in common. They are also both hunting for the same investors. Two recruiters at a developer conference share every interest and want exactly the same people. Similarity tells you two people could hold a conversation. Goals tell you whether either of them will get anything out of it.
Similar versus complementary
A similar match is two people who look alike on paper. A complementary match is two people whose needs fit together: one has what the other is looking for.
Both have a place. Peer conversations are valuable, especially for someone new to a field who wants to compare notes. But the meetings people travel for are usually complementary. The founder wants an investor. The buyer wants a supplier. The company that is hiring wants candidates, and the junior engineer wants someone who has done the job for ten years.
If your matchmaking only looks at overlap, it will keep surfacing peers and competitors, because they are the people who look most like you.
Why topic overlap runs out
Shared interests still matter. They give two complementary people something specific to talk about. An investor and a founder who both care about battery storage will have a better first 20 minutes than an investor and a founder in unrelated fields.
But overlap stops helping after a point. Sharing two or three specific interests is meaningful. Sharing eleven usually means both people ticked most of the list. Good scoring caps the reward for overlap so it can't drown out everything else.
In Conecta, shared interests count for up to 40 points and stop adding at five shared interests, while goal compatibility counts for up to another 40 on its own. That means a pair with three shared topics and well-matched goals can rank above a pair with ten shared topics and clashing goals, which is usually the right call. The matchmaking page explains the rest of the scoring.
Goals belong to topics, not to people
A person rarely has a single goal. A founder at a startup event may want investment for the company and also need to hire two engineers. On a topic they know well, they may be glad to mentor someone newer. Each of those goals attaches to a different topic.
That is why goals work best when they are attached to interests. The attendee picks an interest, then says what they want from it. In Conecta the options are Networking, Trading, Recruiting, Investing and Mentoring.
Most of these describe one side of a relationship:
- Investing is the person with capital to put into companies.
- Recruiting is the person with jobs to fill.
- Mentoring is the person offering experience.
Each pairs best with people on the same topic who chose Networking: the founders, candidates and mentees who are open to those conversations. Trading is different. Buyers and sellers both choose it, so Trading pairs best with Trading.
Four pairings that work
Founder and investor
The investor picks Investing on "Fintech" and "Fundraising". The founder picks Networking on the same interests. That is one of the strongest pairings a scoring system can produce, and it should rank near the top of both lists.
Two investors who both chose Investing rank low. At a pitch event they are usually looking at the same deals, and they will find each other at the bar without help. For the setup side, our page on startup and investor events shows how this plays out alongside Demo Day.
Recruiter and engineer
The recruiter picks Recruiting on "Backend engineering". Engineers who chose Networking on the same interest appear on the recruiter's list. Two recruiters, both on Recruiting, rank low, since they are competing for the same candidates.
If your audience is wary of being recruited at every turn, limit which interests allow Recruiting at all, for example only a "Careers" category. That keeps hiring conversations available without turning every engineering topic into a job board.
Mentor and mentee
The mentor picks Mentoring on "Product" or "Go-to-market". Founders and early-career attendees who picked Networking on those interests see the mentor near the top. This pairing is the backbone of accelerator demo days and student tracks, where the point of the event is for experienced people to give time to newer ones.
Buyer and seller
At a trade fair, both sides choose Trading. Here the topic overlap matters more than anywhere else, because a buyer of packaging has no use for a seller of software, however well their goals fit. Trading with Trading on the same product category is the strongest signal you will get, and it is why matchmaking at trade fairs leans so heavily on a well-built category list.
Pitfalls
Everyone picks everything
This is the most common failure. Given a long list, many attendees tick most of it "just in case", and some choose the most ambitious-sounding goal on every interest. The interest part of the score becomes the same for everyone, so it stops helping anyone rank.
The fixes are in the setup:
- Keep the interest list short, so there is less to tick.
- Restrict which goals each interest allows. If "Product" only offers Networking and Mentoring, nobody can choose Investing on it.
- Describe each goal in plain words during onboarding. "Investing: you put money into companies" stops founders choosing it because it sounds important.
Our guide to onboarding questions for matchmaking covers building the list in detail.
Goals that don't fit the event
Only offer goals that make sense for your audience. A trade fair needs Trading and Networking. A startup event needs Investing, Mentoring and Recruiting next to Networking. If a goal has no counterpart in the room, people who choose it will see empty lists and conclude the matching is broken.
You control this through which goals each interest allows. If no interest allows Trading, nobody can pick it.
Hiding the reasoning
People act on recommendations they understand. A bare percentage next to a name invites the question "based on what?". In Conecta, each profile carries a "Why you should meet" note that shows the shared interests, "Your goals align" where goals fit, and "Similar to people you saved" where the AI layer has learned from someone's saved profiles. An attendee can see in a few seconds whether the suggestion makes sense to them.
Expecting AI to fix bad inputs
AI matching that reads job titles, companies and profile descriptions, and learns from saves, accepted meetings and chats, can improve rankings as the event goes on. People someone keeps skipping drop down their list. What it can't do is invent goals nobody captured. If the interest list is vague and every goal is allowed everywhere, the AI is ranking noise.
Treat it as a layer on top of clear goals, and keep it optional. Organizers who want a fully explainable list can switch it off.
Explaining goals to attendees
Attendees will only choose goals well if they understand them. A short line under each option in onboarding does most of the work. Wording you can adapt:
- Networking: you want to meet people working on this topic, including investors, employers and mentors who are interested in it.
- Trading: you buy or sell products or services in this area.
- Recruiting: you are hiring for roles in this area.
- Investing: you put money into companies in this area.
- Mentoring: you are offering advice to people earlier in their career or company.
Then say plainly that Networking is the right choice for most people on most topics. A founder looking for investment should pick Networking on "Fundraising", not Investing, because that is what puts them in front of the people who chose Investing. Skip that sentence and some founders will pick Investing because it sounds closer to what they want, then wonder why the investors, who also chose Investing, sit so low on their list.
Setting goals by event type
| Event type | Goals to allow | Strongest pairing to aim for |
|---|---|---|
| Startup and investor event | Investing, Mentoring, Recruiting, Networking | Investor and founder |
| Trade fair | Trading, Networking | Buyer and seller |
| Technology conference | Recruiting, Mentoring, Networking | Recruiter and engineer |
| Industry association meeting | Networking, Trading | Members and suppliers |
Start from the pairing you most want to produce, then work back to which goals each interest should allow.
Putting it together
Shared topics tell you two people could talk. Goals tell you whether the conversation will be worth having. Build matching that rewards complementary goals first and uses topic overlap to make those meetings specific, with a cap so overlap can't swamp everything else. Keep the interest list short and restrict goals per interest. Then show attendees why each person was suggested. The matchmaking features are built around exactly that order of priorities.