How to create an algorithm to create offline meetings
With Wildcard we made a calculation from the start of our journey to match people based on interest, to create a spark between people so both parties have an icebreaker into a conversation.
No discovery tools have cracked this as well as Tinder has, and to my opinion ever. To connect you with the right people at the places you visit is a huge challenge.
There have been many developments over the last 25 years on communication technology, although they were mainly aimed to go online to online.
The online to offline space is where we are with Wildcard, and it’s lonely out there.
I’m not talking about search, you can search for Ultimate Frisbee in a town you visit and you can easily find people throw a disc with. That’s really specific and caters only to you if you have a concrete and active request.
I’m talking about serendipity and discovery — with the consequence of mystery and excitement. You don’t like to search for a partner just as much you don’t want to search for your new conversations you have with the people around you.
There’s a magic around discovery, it’s almost like a game, a mystery, and people like to be teased now and then
So we looked at Tinder and based our matchmaking algorithm loosely on theirs.
Tinder works something like this
The people you see on your Stack, are calculated by a score, that could have (we don’t know for sure) these parameters:
- Popularity calculated by the amount of users liking you (swiping right),
- The person or amount users super-liking you, if any exist,
- Newness of account, new users are positively discriminated so they will stick to using the app,
- Newness of account to a location, matching with people at new locations is sort of the apps unique selling point,
- Some public data from your Facebook profile- how many friends you have, the likes you share with a person, etc,
- Your conversational frequency once matched,
- Reports on you, if any exist.
That’s all good, we assumed this is what it’s all about at Tinder, they must have more parameters but that is their undisclosed ‘secret sauce’.
Since we’re not a dating app, but an ‘online-to-offline-meeting-interesting-people-app’ ha ha, we added some more parameters.
Wildcard matchmaking calculates even more
- The amount of Twitter followers you have, adds to the popularity score,
- The proximity of you relative to the swiping user, the closer the better,
- Matching words in your bio and goal (If a user used ‘design’ in her bio, you would be matched quicker if you had ‘design’ in your goal),
- Last activity, the more activity the higher up the Stack,
- Last update of goal, this one is big, (and we even experimented with sending out a notification if your match changed its goal, if you haven’t started talking yet).
These scores are adjusted every time, and for our customers we even make custom Stack algorithms. We like to experiment with this a lot, even with AI learning algorithms. Because every campus and place is different; one has more events, the other more workspaces, and different buildings influence the calculations.
We believe it’s key in current times to have more offline meetings. We wrote about this last week as well:
We live in an online world, where screen time exceeds offline face-to-face time. We need to step outside of our comfort…words.wildcard.works
An offline meeting is worth a thousand texts
This is true because communication is more than words, more than spoken words, more than just looking at each other.
Communication is a dance of sending and receiving signs and sentences, together with one person or many.
Offline conversations are so much more alive than the ones behind a screen. It’s easier to bounce ideas back and forth, find icebreakers and really get each other’s perspective (and hugs).
Do you have any ideas to improve our #TinderforBrains? Start the conversation with us. Come visit us in Amsterdam #MyHQisTQ
for a 🍺 or ☕️. Thanks (and thanks for pressing that 💜)!