What Behavioural Research Quietly Teaches You About Desire
I used to think desire was the one part of life that refused to sit still long enough to be measured. Then I started spending my days with behavioural data, and the joke turned out to be on me. People are astonishingly consistent once you stop listening only to what they announce and start watching what they do with their time.
That is the first quiet lesson. Desire leaves a trail.
Not a cinematic one. Not rose petals and a well timed text that arrives while a song is playing. More often it is a pause that lasts two hours instead of two minutes. A profile opened and closed. A message drafted and never sent. A body that leans in during a conversation and then finds a reason to check the fire exit. None of that is poetry. It is behaviour. And behaviour, unlike a speech about being “bad at texting,” has the decency to repeat itself.
In research we talk about stated preference and revealed preference as if they were two polite colleagues who sometimes disagree. In actual life they have a messy affair and then pretend not to know each other at breakfast. Someone will tell you they want something slow, thoughtful, no games. Then they will only light up for the person who keeps them slightly unsure. Someone will say they are done with intensity. Then they will replay a three week thing for months because it felt like weather. I am not mocking that. I have done both. The point is not that people lie. The point is that the mouth is often still writing the press release while the rest of the system has already voted.
Once you accept that, work stops being abstract. A click is not just a click. It is attention with a timestamp. A delay is not just rudeness. It is a ranking of priorities that the person may not even have admitted to themselves. A small avoidance, the cancelled plan that always happens on the same kind of night, is often more honest than a paragraph about needing space.
I should say this early, before anyone decides I have become the office cynic with a spreadsheet where a heart should be. Data does not make desire cheap. It makes it specific. That is a kinder thing than mystery, which is frequently just poor communication in a better jumper.
Take latency. Response time is one of those variables people treat as etiquette when it is often motivation wearing a watch. If someone can answer a work Slack in four minutes and your message in four days, you do not need a séance. You need a sense of proportion. The nervous system is not confused. It is allocating. I say that as someone who has stared at a grey tick and built an entire attachment seminar in her head. Anxious-secure leaning, if we are being precise, which I like to be. I can name the spiral and still step in it. The research does not cure that. It just stops me calling the spiral evidence.
There is a version of this that becomes cruel, and I have watched it happen in labs, in product meetings, in friendships. You start treating people like slot machines. Variable reward. Intermittent reinforcement. The person who texts just often enough to keep the hope slightly fizzy. Behavioural science has language for that, and the language is useful until it turns you into someone who cannot be surprised anymore. I refuse that ending. The useful reading is not “everyone is manipulating everyone.” The useful reading is “attention is finite, and desire spends it.”
Clicks taught me that more brutally than any date. In digital behaviour you see the gap between what people save and what they return to. They will favourite the worthy article. They will reopen the thing that made their stomach dip. They will hover over a name. They will search a person they claim not to think about, then close the tab as if the tab were the problem. I have sat with heat maps that looked, uncomfortably, like longing. Clusters of attention. Dead zones. The same three points of contact, again and again, like a hand that knows the route even when the owner is busy being sensible.
It is almost intimate, that kind of work, which is a strange sentence to write about datasets. You are not in the room. You do not have their face. You have traces. And traces are oddly tender if you let them be. Somebody wanted, then flinched. Somebody approached, then found a reason to tidy the kitchen. Somebody kept a conversation alive past the point of convenience. Convenience is underrated as a measure of feeling. People make time for what they cannot quite leave alone.
Avoidance is the part everyone wants to dress up. We call it boundaries when it is fear. We call it intuition when it is a bias. We call it “I just was not in the right headspace” when the headspace in question was simply not that interested. Sometimes those phrases are true. I am not in the business of calling every soft exit a crime. I like people who can say no without turning it into theatre. What the research keeps showing, though, is that repeated avoidance is information. One cancelled Tuesday is weather. Four cancelled Tuesdays is climate.
There is a dry little joke I make in my own notes, never in a client meeting. The body is a terrible liar and a gifted administrator. It will keep the calendar honest even when the mouth is writing fiction.
Desire, in the behavioural sense, is approach. Not the speech about approach. The movement toward. You can feel attraction and still not approach. That is not a plot hole. That is conflict. Approach and avoidance can fire together, which is why a person can want you and go quiet in the same afternoon. Ambivalence looks like mystery from the outside. Up close it looks like two systems arguing and neither of them unionised.
I think about this a lot because my job sits on both of my degrees. The maths wants base rates and clean variables. The psychology wants context, attachment history, the fact that a delayed text lands differently if you grew up listening for footsteps on the stairs. You cannot put a person in a model and delete the childhood. You also cannot use the childhood as an excuse for never choosing. That is where I get impatient, in a warm way. Pattern is not prison. Pattern is a weather report. You still decide whether to go out.
If you want the unembarrassed bit, here it is. Sexual desire follows a lot of the same rules as any other motivated behaviour, which sounds unsexy until you notice how sexy competence becomes once you understand that. People get wet for attention that stays on their face. They get bored by performance that never checks whether anyone else is in the room. They return to the partner who can change pace. They disappear on the partner who treats them like a mood they can pick up and put down. Variety can be delicious. So can repetition, if the repetition includes being recognised. I have never found those two things as opposed as the internet pretends.
Behavioural research will not tell you who to sleep with. It will tell you that novelty spikes, then settles, and that the settling is not a failure unless you needed the spike to feel alive. It will tell you that effort changes value. We like what we have worked for, sometimes more than what arrived easily, which is either romantic or a warning label, depending on the week. It will tell you that people underestimate future feeling when they are calm, then overestimate their own restraint when they are not. The hot cold gap is not a personality flaw. It is a design feature of being an animal with a diary.
I lost my virginity at eighteen in a way that was clumsy and kind, and I still remember the kindness more than the choreography. That is data too, just not the sort you put in a dashboard. Later there were louder nights, different rooms, the first time I asked for exactly what I wanted and did not apologise for the sentence. What stayed was never the notch. What stayed was the person who watched my face as if it contained information. That is revealed preference again. I can talk about adventure all afternoon. My attention still goes to being met.
There is a particular Essex warmth I will not disown here, because it belongs in this piece. I will compliment a stranger’s jacket. I will talk in a bathroom queue. I like a night out that ends with cold air and the feeling of having spent yourself properly. None of that is the opposite of taking desire seriously. The stereotype wanted loud and disposable. I kept the pleasure and threw away the punchline. Behavioural work actually fits that better than people think. You can enjoy the surface and still read the structure underneath. You can dance and still know when someone is only dancing toward the next available body.
What the job taught me, more than any paper, is how often desire is just ranking. Not ranking people like objects on a shelf. Ranking states. Comfort versus charge. Certainty versus the little lift of not knowing. Being chosen versus being free to leave. Most of us want a cocktail of those and then act shocked when the mix is unstable. I do it. You do it. The person you are waiting on is doing it while making tea.
Cynicism arrives when you notice the ranking and decide it means nobody means anything. That is a failure of interpretation, not a triumph of intelligence. Of course people are inconsistent. They are running old software on new nights. They are tired. They are half in love with a version of themselves that only appears when someone new is watching. They are scared of wanting a thing they might not get. Fear looks like indifference if you only measure output and ignore the cost of trying.
So I try, at work and outside it, to read the small avoidances without turning them into a prosecution. A person who never initiates is telling you something. A person who initiates only after you go quiet is telling you something else. A person who remembers the tiny detail from three conversations ago is not performing magic. They encoded you. Encoding is one of the more flattering things a nervous system can do. I notice it immediately. I also notice when I am the only one doing it. That is not me being needy. That is an imbalance in investment, and investment is visible if you are willing to look without decorating it.
Digital life made all of this louder, which is why clicks belong in an essay about desire even if that sounds unromantic. We used to have to walk across a park at two in the morning because neither person wanted the conversation to stop. Now we have read receipts and last seen and the little betrayal of a like on a photo from fourteen weeks ago. The tools changed. The underlying behaviour did not. Approach still looks like extra steps. Avoidance still looks like a sudden love of sleep, work, and “this week is mad.” I have had weeks that were actually mad. I have also used madness as a velvet rope. I try to admit the second one at least to myself. Emotional honesty is one of my non negotiables, and it would be cheap to demand it from other people if I will not practise it in the dataset of my own excuses.
There is a maths comfort in all this that I will not pretend I do not enjoy. Base rates keep you from turning one beautiful night into a forecast. One intense weekend is a point. Several weekends with the same shape are a line. People love points. They make speeches out of points. I prefer lines. Lines are less glamorous and much harder to argue with. If someone lights up in person and vanishes in text, that is a line. If someone is average in a room and extraordinary in the follow through, that is a better line than fireworks. Follow through is an underrated erotic quality. I will die on that hill, politely.
Dry humour helps, or I would take all of this far too seriously and start counting tiles before breakfast. The tiles thing is real, by the way. When I am anxious I count. Steps. Beats. The number of times a thought returns. It is not mystical. It is a brain reaching for structure. Behavioural work is that habit with a salary. You count so the feeling does not get to be the only narrator.
The intimacy of the job, if I am going to honour the brief, is this. You begin to see that desire is rarely a bolt of lightning and often a pattern of return. People come back to what regulates them, or what dysregulates them in a way they have learned to call chemistry. Distinguishing those two is the whole art. Chemistry that costs you sleep, appetite, and your ability to finish a sentence can still be real. It is just not always good for you. Compatibility looks dull in a headline and extraordinary in a kitchen at half ten on a Sunday, when nobody is performing and somebody still wants your mind.
I think about Sunday lunches more than a serious researcher is supposed to. Arguments about politics, then a record, then the particular relief of people who did not leave because it got loud. That is desire too, just stretched over years. The modern conversation is obsessed with the spark and undereducated about the stay. Behavioural research, if you let it, corrects that. Repeated voluntary contact is one of the least poetic and most reliable signals we have. You do not keep choosing a person by accident. Accidents do not have that kind of attendance record.
None of this means you should become a detective. Please do not. Surveillance is not intimacy, and I say that as someone professionally adjacent to traces. The point is not to build a case file on someone you fancy. The point is to stop gaslighting yourself with their adjectives when their verbs are doing something else. Listen to the verbs. Verbs are shy, but they show up.
If there is a positive ending here, and I want one, it is that reading behaviour can make you gentler rather than harder. You stop taking every delay as a verdict on your worth. Sometimes it is a verdict on their capacity. Those are different sentences. You stop chasing the person whose attention is a drip feed and calling it passion. You start noticing the one who answers in a way that matches their life, not their brand. You get better at asking for what you want because you have seen, over and over, that unclear desire produces messy data. Clear desire produces cleaner rooms. Not perfect rooms. Cleaner.
I still want the walk across the park. I still want the conversation that refuses to end. I still want sex that includes a mind, a face, and the kind of aftercare where somebody talks instead of collapsing into a performance of being cool. None of that is less likely because I understand reinforcement schedules. If anything it is more likely. You cannot choose well if you keep calling confusion a vibe.
So yes. People tell you who they are with delays, clicks, and small avoidances. Once you can read that without becoming cynical, the work stops feeling like a pile of numbers and starts feeling like what it always was. Attention, fear, appetite, hope, the little human decision to move closer or step aside. Desire is not above measurement. It is just offended by it.
I measure it anyway. Then I go home to Essex when I can, put on a dress that makes me feel like myself, and try to live as if the data were a lantern rather than a weapon. That is the whole trick. See clearly. Stay warm. Leave when the pattern says this is not mutual. Stay when the pattern says somebody keeps coming back, not because they are hooked on the uncertainty, but because they like the actual person standing in front of them.
That last one is rarer than the films suggest and more common than the cynics allow. I would bet on it. Carefully. With a sample size bigger than one good night.


