How does MeAndWho Club choose introductions?
The rules behind every introduction, in plain words. Scoring version score-v1.
In short
MeAndWho Club matches with fixed rules over each member’s own answers, not with the AI. Two people must pass each other’s filters in both directions, their fit is scored both ways across five areas and ranked by the less keen side, and each must be in the other’s top 5 (currently). During our pilot a person checks every suggested pair before anyone is introduced, and a chat opens only if both say yes.
Last checked
Step 1: who can be introduced at all?
Only verified, active members, and only pairs that pass each other’s filters both ways.
Read the details
Before any scoring, a member must be in the pool: phone checked, live selfie check passed, at least two approved photos with one matching their selfie, required consents given, and not paused. Then three filters run in both directions:
- Who you want to meet. You are each someone the other wants to meet. This works for every gender and orientation.
- Age range. Each of you is inside the other’s chosen age range.
- Dealbreakers. Each of you meets the other’s dealbreakers. You can set at most three, because every dealbreaker shrinks the pool.
People who have blocked each other are never paired, and a pair that has been suggested once is never suggested again.
Step 2: how is fit scored?
Across five areas, from your answers to the questions, weighted by what you said matters.
Read the details
| Area | What it looks at |
|---|---|
| Intentions | The kind of relationship you want. This carries the most weight. |
| Values | Children, family living arrangements, how you balance career and life, and religion or community only if you said they matter |
| Lifestyle | Diet, drinking, smoking, education, languages you share, and interests in common |
| How you connect | How you communicate, how much time you like together, social energy, and how you handle disagreement |
| Location | Your area of Bengaluru, if you both answered: by the areas you said you’d travel to or, if you didn’t say, by how similar your answers are |
Within each area, each question counts in one of four ways:
- Your stated preference. Where you said what you want in a partner, it counts by how much it matters to you: “Doesn’t matter to me” not at all, “Nice to have” half, “Important” and “Dealbreaker” fully.
- No stated preference. Where you didn’t say what you want in a partner, similar answers still count, at half weight. Religion and caste are the exception: without a stated preference they don’t count at all.
- Similar answers. For things like how you communicate, the same answer counts fully and a neighbouring one counts half.
- Any shared language counts as a fit.
- Interests count by the share you have in common.
Religion and caste or community only count where a member has said they matter to them. If someone makes one a dealbreaker (“Must be the same”), people who didn’t answer it aren’t introduced to them. These answers are never shown on an introduction card, and your answers to them are never sent to the AI.
Step 3: why is fit scored in both directions?
Because an introduction only works if it suits both people.
Read the details
Each pair gets two scores: how well the other person fits what you want, and how well you fit what they want. The two are combined so that a weak side pulls the score down more than a simple average would, and pairs are ranked by the weaker of the two.
An illustration
With made-up numbers on a scale of 0 to 1: a pair that fits 0.8 one way and 0.5 the other ranks by 0.5. A pair that fits 0.9 one way and 0.4 the other ranks lower, at 0.4, even though one side is keener. A lukewarm side pulls a pair down, however keen the other side is.
A missing answer lowers how much that area counts for the pair, not the fit itself, so skipping an optional question doesn’t count against you.
Step 4: which pairs are suggested?
Only pairs that clear a minimum score and are each in the other’s top 5.
Read the details
- The combined score must clear a minimum. Below it, no introduction, however few people there are.
- Each person must be in the other’s top 5 of eligible people (currently).
- Each member has at most 2 introductions waiting for an answer and at most 3 new ones a week. These are ceilings, not targets: there is no fixed schedule.
Step 5: who checks a pair, and what does the AI add?
A person reviews every suggested pair, and the AI writes the reason you see.
Read the details
The reviewer sees both introduction cards, the notes each member confirmed for matching, the reason the AI wrote and the score breakdown, never raw answers, AI conversations or verification data, and every look is logged. Only an approved pair becomes an introduction.
The AI writes the “Why we suggested this” note from what you both chose to share on your cards. It may not mention scores or percentages, and it never promises that a pair will work out. Notes from your AI conversation help the reviewer understand you; they don’t change the score.
Later, once quality measures hold for each group of members, well-understood pairs may be sent without a person checking each one, and the system falls back to human review automatically if a measure slips. No stage can skip double opt-in: a chat opens only when you both say yes.
Questions about the rules
Does the AI decide my match score?
No. The score is plain arithmetic over your own answers and preferences, using fixed rules that are the same for everyone. The AI writes the short reason shown on an introduction, and during our pilot a person reviews each suggested pair, but the AI does not calculate who fits whom.
Will skipping a question hurt my chances?
No. A missing answer lowers how much that area counts for a pair; it is not treated as a mismatch. The one exception is a dealbreaker: if someone made a question a dealbreaker and you left it unanswered, you won’t be introduced to them, because we can’t confirm you fit it.
Do popular people get more introductions?
Not by design. Pairs are ranked by the less keen side of the fit, not by how many people would like someone, and each member can have at most 2 introductions waiting and 3 new ones a week. We also measure, inside the team, how introductions are spread across members.
Can the rules change?
Yes. The weights, the minimum score and the caps are settings, chosen as sensible starting values and reviewed once there is real feedback. Every change needs a written reason and is logged. When a rule described here changes, we update this page and its date.