Signs A New Relationship Has Real Potential

What can and cannot be predicted

A warning before any list of signs, including this one and the others on this site.

The longitudinal research on relationships is largely conducted with established couples, often married, often followed for years. Applying its findings to a relationship that is four months old involves an inferential leap that most articles make silently.

The honest position is that early prediction is weak. What the research supports is which behaviours and patterns are associated with better trajectories in couples who are already together — and it is reasonable to treat those as a rough guide to what to look for early, while remembering that the evidence was not collected there.

So this article is organised by strength of evidence rather than by how compelling the sign sounds. The first section has real longitudinal support. The last section is folklore that circulates as fact. One variable does escape the caution above, because it has been measured on exactly this population — unmarried couples, both partners reporting, followed over two years — and that is whether the two people hold the same level of commitment: what asymmetric commitment predicts is the closest thing here to an early marker with evidence collected in the right place.

The same caution applies to any number a website can hand you, our own relationship satisfaction test included. It describes where things stand now from one person’s side, and it says so on the page rather than dressing a snapshot up as a forecast.

Sign: they respond when you reach for them

The single most transferable finding from the observational work is about responses to small bids for attention — the moment one person says something minor and the other either engages or does not.

Gottman's lab built a large part of its reputation on this, and the widely quoted percentages — that couples still together turned toward 86% of the time and those who divorced 33% — turn out not to appear in the peer-reviewed papers reporting that coding system, as the review of what those studies actually measured sets out. What those papers do show is that the rate of turning toward in ordinary moments was associated with how a couple behaved during conflict weeks later. It is a frequency measure over ordinary time, which is exactly what makes it usable early: you do not need years of history to observe it, only a few weeks of ordinary contact.

In a new relationship this is easier to see than in an old one, because nothing has become automatic yet. Watch what happens when you say something inconsequential. Does it get received?

The caution is that early relationships are on best behaviour, so a high rate at six weeks is less informative than a maintained rate at six months. What is informative early is a low rate, since that is the effortful direction to fake.

Sign: your good news is interesting to them

The capitalization research is the finding I would put second, and it is underused because everyone is watching how a partner handles bad news.

A dyadic study using intensive longitudinal methods found that capitalizing on everyday positive events — sharing them with a responsive partner — uniquely predicted daily intimacy and well-being, in a sample of couples where one partner was coping with breast cancer. That population matters for the reading: these were people with no shortage of hard news, and the sharing of good news still did independent work.

What to look for is not congratulation. It is interest. Questions about the thing. Bringing it up unprompted two days later.

The failure mode is the flat receipt — that's great — which acknowledges without engaging. It is easy to miss because nothing is technically wrong, and over months it is one of the more corrosive absences.

Sign: conflict happens and does not escalate the same way twice

A new relationship with no conflict at all is not evidence of compatibility. It is usually evidence of insufficient time or of one person absorbing everything.

What is worth watching is the shape of the first few disagreements. Specifically: does either of you offer a way out mid-argument, and does the other take it? The observational research treats these repair attempts as differentiating, and when Gottman, Driver and Tabares coded what newlyweds actually did, the attempts that worked were the early ones — offered in the first three minutes — and the ones that addressed the emotional climate rather than the argument: humour, affection, taking a share of responsibility. Appeals to logic did not do the same work.

So the early sign is not a partner who argues gracefully. It is one who offers something warm early, and one who takes it when it is offered.

The pattern to watch for in the other direction is demand and withdraw — one person pressing, the other retreating, the pressing escalating in response. A diary study of 116 couples found this occurs at equal frequency in both gender directions and predicts negative emotion and lower conflict resolution. It is covered at more length in the demand and withdraw research, including who does which half, and if you want to see which half you tend to take, the conflict style test scores pursuing and withdrawing as separate dimensions — with the caveat, stated on its own page, that the role usually depends on whose issue is being discussed.

Sign: being with them makes your world bigger, not smaller

The self-expansion model describes close relationships as a process in which each person's sense of self grows to include the other — resources, perspectives and identities becoming partly shared.

A prospective study of 147 first-year college students, built on this model and on broaden-and-build theory, tracked daily emotions for 28 days and measured self-other overlap with a new roommate. Students reporting more positive emotions in their first weeks reported greater overlap, and the association held with extraversion and negative emotions controlled (β = 0.25); the ratio of positive to negative daily emotion then predicted overlap a month later even with the earlier overlap controlled (β = 0.17, p = .05). Those are randomly assigned roommates rather than romantic partners, which is the limit worth carrying. What it supports is narrow and still useful: the emotional weather of the time two people spend together shapes how much of each other they take on. The wider evidence on that process is in what builds attraction over time.

What that looks like in practice at four months is fairly concrete. Are you doing things you would not otherwise have done? Have you met people you would not have met? Is there some capability or interest you now have because of them?

The inverse is the sign to take seriously: if the relationship has contracted your world — fewer friendships, dropped interests, less contact with family — that is a pattern worth examining, and it is worth distinguishing between a contraction you chose and one that was produced by disapproval or friction every time you tried to leave.

Sign: the account of their past includes them in it

This one is cheap to observe and surprisingly informative early.

Listen to how a new partner describes previous relationships. Not whether they are kind about their exes — some people genuinely behaved badly and saying so is honest — but whether any of the accounts include a part they played.

A history in which every former partner was crazy, cold or dishonest, and the narrator was consistently reasonable, has a missing variable. It is not proof of anything and it is a strong prior, because the one person present in all of those relationships is the one telling you the story.

The opposite extreme is informative too. Someone who takes complete blame for everything is usually describing a self-punishment pattern rather than accountability, and that has its own consequences.

What you want is texture: here is what they did, here is what I did, here is what I would do differently now.

What the evidence does not support, despite the confidence

Now the folklore, all of which circulates as fact.

Intense early chemistry. There is no good evidence that early intensity predicts duration, and the attachment literature gives reason for suspicion: anxious activation is routinely experienced as chemistry, and it is not a marker of a secure bond. That confusion is unpacked in love or anxious attachment.

Body language reading. The popular claims rest on misreadings — the 7-38-55 figure being the standing example — and what survives meta-analysis is much narrower. See what body language actually predicts.

Flirting as proof of interest. People flirt for fun, for a lift in self-esteem and to get things done at work, as well as for courtship, and in recorded conversations some people become more reserved rather than warmer when they are attracted. Why flirting proves less than it seems sets out that evidence.

Response timing. The three-day rule and its relatives are folklore with a confident delivery, examined in the 72-hour rule in dating.

Milestone months. The idea that a relationship declares itself on a schedule — at three months, at six, at nine — has nothing measured under it, and the best dataset on when dating relationships actually end shows a spread from one month to five years with no clustering at any of those points: the 3-6-9 rule in dating.

Shared hobbies. Similarity in values and life goals has support. Liking the same films does not, and the two get conflated constantly.

Never fighting. Discussed above. Usually a measure of elapsed time or of asymmetric accommodation.

The one thing worth doing rather than watching

All of the above is observation, and observation has a limit that people reach much later than they should.

At some point the useful move stops being pattern-collection and becomes a direct conversation about what each of you wants. Not a demand for definition at week three, which is a different error. But the tendency to keep gathering evidence indefinitely is usually avoidance wearing the clothes of prudence.

A relationship's potential is not only a property of the other person. It is also determined by whether the two of you can have a straightforward conversation about where it is going without either of you managing the other's reaction.

That conversation is itself the best available test, and no amount of sign-reading substitutes for it. If you want a structured way in, questions that reveal real compatibility is built for exactly this stage.

And the line that is not about potential

One category does not belong on a spectrum from less to more promising.

Early monitoring — wanting to know where you are, who you are with, reading your messages. Escalating jealousy framed as care. Pressure to reduce contact with friends or family. Being made responsible for managing their anger. Any of these appearing early is not a sign of intensity or of a relationship that needs work.

These patterns tend to establish themselves gradually and to be explained, at each step, as something else. That is the reason to name them explicitly rather than trusting that they would be obvious.

If this describes what you are reading about, the useful next step is a domestic abuse service rather than an article about signs. In the UK, the National Domestic Abuse Helpline runs on 0808 2000 247; in the US, the National Domestic Violence Hotline is 1-800-799-7233. Both are free and confidential.