The Anti-Interview
Why questions kill attraction, and what to do instead. The four moves that replace the interview stack.
Want this kind of read at your fingertips while you text?
Install DateIQ — the AI wingman pulls from this library when it can help.
There is one anti-pattern that kills more threads than every other failure combined. It's the interview stack: a string of questions, often well-meaning, that turns a flirtation into a hiring process. "What do you do? Where are you from? Any siblings? What do you do for fun?" Each question is fine on its own. Stacked, they put you on the wrong side of the dynamic — the one asking permission to keep talking. The interviewer is the submissive role. The interviewer is also boring.
This is the single hardest pattern to break, because asking questions feels like the right thing to do. You were told it was polite. You were told women like men who are interested. Both are partly true. Neither survives contact with the actual mechanics of attraction over text.
Why questions kill
Three reasons.
First, questions transfer effort from you to her. Every question you ask makes her do the work of answering. A thread is interesting when both parties are bringing material to the table. When you ask questions, you stop bringing material — you're just asking her to bring more.
Second, questions reveal that you have nothing of your own to say about the topic. If you ask "what kind of music do you like?" you're announcing that you have no opening opinion of your own. The same question asked as "I'm guessing you have weirdly specific taste in 2000s indie based on your bio — am I close?" is a totally different move. One reveals a void; the other reveals a person.
Third, questions create a job-interview rhythm: you ask, she answers, you respond with "that's cool," you ask again. That rhythm has no tension. No tension, no attraction.
Here's what it sounds like when the interview pattern takes over. Notice how every "me" line is a question and every "them" line is doing the work of answering:
Twelve messages in and you've learned six facts and built zero tension. She is, technically, replying.
What to do instead
There are four moves that do the work of a question without being one. Use them in rotation.
1. Assumption
Instead of asking, guess. "Let me guess — middle child, ran cross-country in high school, has strong opinions about brunch." She'll confirm or correct. Either way, you've shown personality. The assumption is the seed of the Let Me Guess framework — see that article for the full play.
2. Inference
Read the cues she's giving and reflect them back as a statement. She mentioned her dog three times in five messages: "I'm gathering Luna is essentially a third party in this conversation." That's not a question — it's an observation that invites her to expand. Inferences land warmer than questions because they show you've been paying attention.
3. Statement
Just say the thing. If she mentions a band, don't ask if she likes their last album — say what you think of it. "Their last album is genuinely a step down and I'll fight anyone who says otherwise." Statements give her something to react to. Reactions are more textured than answers.
4. Observation + pivot
Notice something — about her, about the conversation, about the topic — and pivot from it to a broader thread. This is essentially the structure of O.C.A.T. at the message level. Observation does the work; pivot opens a door without locking her into a one-line answer.
When questions are actually fine
Questions aren't poison; stacks of questions are. One specific, charged, open-ended question dropped into a thread otherwise made of statements and inferences is fine — and in some cases necessary, especially if you're trying to move toward logistics. The rule of thumb: if your last three messages have been questions, your next message must not be one.
A second exception: when warmth is unmistakably high, questions read as genuine interest rather than as desperation for material. A woman in the attraction stage who asks you personal questions is communicating investment, and you can do the same back. The interview-stack risk is heaviest in early chat and mid chat, when the dynamic is still being set.
The translation table
For the next ten threads you have, run every question you draft through this filter:
- "What do you do?" becomes a guess: "Let me guess, something in design with a job title that nobody outside the company understands."
- "Where are you from?" becomes an inference: "You don't sound like someone who grew up in the city. Smaller town, somewhere with cornfields nearby — am I close?"
- "Any siblings?" becomes an absurd reframe: "Reading you as either an only child or the eldest of four. No middle ground."
- "What do you do for fun?" becomes a statement: "I've been telling myself I'm going to start climbing again for two months and have so far been to the gym once. Your move."
Each translation does more work in fewer words. None of them ask permission. None of them put you in the interviewer's seat. Do this for a month and your threads will look different.
Same opener, run through the four moves above. Watch what changes — the rhythm, the energy, who's bringing material:
Same length, same person, completely different gravity. She's volleying instead of answering.
Pairs well with
Let Me Guess
An inference framed as an absurd assumption. Invites her to correct or co-sign — always lighter than a bare question.
O.C.A.T.
A four-beat opener recipe: Observation, Comment, Assumption, Topic. Anchors a cold message on a specific detail and turns it into a thread.
Assumption Close
Propose the date as a confident statement, not a question. Concrete time, place, and activity — never invent specifics.
Want this kind of read at your fingertips while you text?
Install DateIQ — the AI wingman pulls from this library when it can help.
Want this kind of read at your fingertips while you text?
Install DateIQ — the AI wingman pulls from this library when it can help.