Red Flags vs. Real Flags: A Founder's Guide to What's Actually Worth Worrying About
Somewhere in the last few years, "red flag" stopped meaning "warning sign of harm" and started meaning "mildly annoying trait I noticed on a second date." Texts in all lowercase: red flag. Orders the same drink every time: red flag. Doesn't like dogs: red flag, apparently a serious one. We've built an entire vocabulary — beige flag, green flag, red flag, the ever-mysterious "ick" — for cataloguing minor personality quirks as if they were structural defects.
Meanwhile, the actual predictors of whether a relationship works or quietly falls apart in eighteen months get almost no airtime, because they're less fun to post about. They don't fit in a fifteen-second video. Decades of relationship science has actually gone looking for these predictors, in labs, with real couples, tracked over real years — and the findings look almost nothing like what trends online. So our founder, who has now sat across the table from thousands of people at the very start of their dating lives, put together a working list. Not the flags the internet is obsessed with. The ones the data — and years of watching people in real rooms — actually backs up.
The Noise: Beige Flags
The "beige flag" trend deserves credit for at least being funny — mild, oddly specific habits people find quietly unsettling. Someone who narrates what they're doing while they're doing it. Someone whose whole personality is one TV show. These are almost never predictive of anything. They're personality texture, not personality problems, and treating them as red flags mostly just trains people to opt out of perfectly good matches over nothing.
The Noise: "Different Communication Styles"
This one sounds serious and almost never is. Someone who takes four hours to text back and someone who replies in ninety seconds are not incompatible — they're just two people who haven't yet agreed on what responsiveness means to each of them, which is a five-minute conversation, not a compatibility crisis. We watch people rule each other out over pacing more than almost anything else, and it's very rarely the thing that would have actually broken the relationship.
The Noise: A Slightly Awkward First Date
An off first date gets treated online as diagnostic — proof of incompatibility, or worse, proof of a personality flaw. In our experience it's mostly proof that two strangers were nervous in a restaurant. Nerves flatten people. The version of someone who shows up relaxed on date four is frequently a completely different, much more interesting person than the one who showed up stiff on date one. Writing someone off after one nervous hour is like judging a chef off the one dish they made under a time limit, live, for a stranger who was also grading their outfit.
The Real Flag: How a Conversation Actually Starts
This is the one piece of internet-famous relationship research that's actually earned its reputation. Psychologist John Gottman's research on married couples found that the way a conflict conversation opens — calm versus harsh — predicted with striking reliability, based on just the first few minutes, how the rest of the conversation and even the relationship would go. It's a small, almost boring-sounding thing: does the conversation start with an attack on the person, or a complaint about the situation. But it compounds. Our matchmakers watch for a version of this even in low-stakes first conversations — not whether someone disagrees with something, but how they open the disagreement. That opening move tends to be a preview of everything that follows.
The Real Flag: How They Talk About Everyone Who Isn't You
Not just exes — the barista, the Uber driver, a difficult coworker, their own family. This is one of the most useful patterns our founder has come to trust, because it's almost impossible to perform for very long. It also lines up with what Gottman's research identified as the single strongest predictor of divorce in his decades of studying couples: contempt, a communication pattern rooted in a sense of superiority over the other person. Gottman's team was able to predict which couples would eventually divorce with well over 90% accuracy largely by tracking contempt and three related patterns — criticism, defensiveness, and stonewalling — often called the Four Horsemen. Someone who's warm to you but sharp or dismissive about everyone else isn't giving you the "good version" of themselves. They're giving you the temporary version. It runs out — and the research suggests that when contempt is the pattern underneath it, it runs out predictably.
The Real Flag: What They Do With an Apology
Not whether they apologize — most people can say the words. The tell is what happens right after. Do they actually adjust, or does the same thing happen again in two weeks with a slightly different excuse attached? Gottman's broader body of work on conflict points to something related here: couples who stay together aren't the ones who avoid conflict, they're the ones who make and accept what he termed repair attempts — genuine efforts to de-escalate and course-correct mid-disagreement. An apology that isn't followed by an actual change isn't a repair attempt. It's a pause button. This is one of the clearest long-term predictors we see, and it's almost never mentioned in a "red flags to watch for" listicle, probably because you can't spot it on a first date — it requires giving someone a second chance and watching what they do with it.
The Real Flag: How They Handle Being Told No
Not a dramatic no — a small one. Plans changing, a boundary being set, someone saying "I'd rather not tonight." The genuinely useful data point is whether that gets met with a shrug or with a subtle campaign to relitigate it. This shows up early, if you're watching for it, and it tends to be one of the more reliable single indicators of how someone will handle disagreement once the relationship gets more serious and the stakes are higher than which restaurant to pick.
The Real Flag: Whether Their Friends Seem to Actually Like Them
Not whether they have friends — whether the friends seem genuinely glad to be around them, versus tolerating them, versus performing loyalty. This turns out to be one of the more well-documented, and most overlooked, predictors in the entire research literature. A large meta-analysis of romantic relationship studies found that when people perceive their friends and family approving of a partner, they're significantly less likely to break up with that partner down the line. Earlier research by sociologists Susan Sprecher and Diane Felmlee, tracking couples over time, found that the more a woman's friends approved of her relationship specifically, the longer that relationship tended to last. Our founder has learned to pay close attention to this in group settings at events, because it's one of the few signals that's almost impossible to fake over a full evening. People can manage a stranger's first impression. It's much harder to manage a decade of friends' body language — and apparently, that body language is telling you something real.
The Actual Point
None of the real flags above are dramatic. None of them make a good viral sound. They require watching someone across more than one interaction, in more than one mood, around more than just you — which is exactly the kind of pattern relationship researchers spent decades in actual labs trying to isolate, and exactly the kind of pattern a single profile or a single date can't surface. That's slower, less satisfying content than "5 red flags to spot in his texting style." It's also the only version of this exercise that's actually backed by anything.
This is, not coincidentally, most of why Luvo exists. You can't extract "how do they treat people who can't do anything for them" from a dating profile, and you can't extract "do their friends actually like them" from three photos and a prompt. You can watch for it across a season of events, across more than one conversation, in more than one room. The internet's red flags are mostly about vibes. The real ones are almost always about power and pattern — who someone is when nobody's grading them for it, and what they do the second time, not just the first. That's a harder thing to screen for. The research suggests it's also the only thing worth screening for.
Sources referenced: John Gottman and the Gottman Institute's research on marital stability and the "Four Horsemen" framework (criticism, contempt, defensiveness, stonewalling); Sprecher & Felmlee's longitudinal research on social network approval and relationship stability; Le, Dove, Agnew, Korn & Mutso's meta-analytic synthesis on predictors of nonmarital relationship dissolution.