Reading someone's profile without guessing

The idea

A profile shows what someone chose to show; it tells you nothing about how interested they’d be in you, so read it as a few facts plus guesses to check.

Sort what you see into three layers.

What it shows. The words in the bio, what is in the photos, the listed interests. “A dog on a beach” is a fact.
What you might infer. “Outdoorsy.” “Well off.” “A parent.” “Her partner.” Each of these is a guess. A guess is worth holding only as a question, never as an answer.
What it can’t show. How interested this person is in you, what they want in someone, and what they are like to talk to.

A photo is one moment that the owner chose to share. A beach photo is not a personality, and a person standing next to someone in a photo does not tell you how they are related.

Two findings bear on the middle layer. In one study, strangers who viewed 100 Facebook profiles estimated the owners’ extraversion, agreeableness and conscientiousness accurately, but not neuroticism 1. The estimates were judged against the owners’ own reports, and the pages were early-2010s Facebook, not dating apps 1. So a page can carry some real information. It does not mean your own first impression is right; treat it as a guess.

A second finding is about what people say they want. Studies of speed dating and early romantic interest, both from one research group, found that self-reported traits and stated preferences, measured before people met, did a poor job of predicting whom a given person would desire 2. What people thought of the partner once they met mattered more 2. The outcome was self-reported desire, not later relationships, and these were not dating profiles 2. It is a reason to hold a list of interests loosely: it says little about who someone will click with, and that includes you.

If you write a first message, make it about something the profile shows. Keep it specific, low pressure, and easy to answer or ignore. Leave out looks, and do not build it on a guess. This is a course principle, not a research finding.

A profile can never tell you how interested someone is in you. When a question asks for that read, “Insufficient information” is the accurate answer.

Why it matters

A guess from a profile can feel like knowledge. A message built on a guess asks someone to answer for a person they aren’t, and a read of interest from a profile has nothing under it.

What it looks like

Example

You come across Priya, 32, on a dating app. Her photos show a pottery wheel and a dog on a beach. Her bio says, “Teaching fifth grade. Ask me about the worst clay mug ever made.” You write: “Was the worst mug yours or a student’s?” The message is about something she shared, and she can answer it in one line or leave it.

The same moment, handled the other way

Counterexample

Same profile. From the beach photo you decide Priya is “spontaneous and outdoorsy” and would want someone who loves dogs, and you open with a line built on that. It fails because it answers a person you made up.

The one thing to remember

The rule

Separate what the profile shows from what you’re guessing, and write only about what it shows.

Name the facts first: the wheel, the dog, the line about the mug. Then pick one to write about. A message gives someone something easy to answer. It cannot make anyone interested.

Try it now

Item 1 of this lesson’s set: A climbing wall and a party photo. Pick what the profile shows first, then the message.

Open the exercise set

In the real world (optional)

If you like, notice once a conclusion you jump to from a single photo, on any page or poster, and what the photo actually showed.

What this doesn’t tell you

The findings are about impressions and stated preferences, not about dating profiles or about any one person. A profile is chosen and partial, and nothing here tells you whether to message anyone. A reply, or none, is the only evidence of interest in talking, and even that is about the conversation, not about the person.

Why do we believe this?

Each numbered mark in the text opens the evidence behind that sentence. Every claim in this course carries one of five labels.

Well supported: A meta-analysis or several independent studies agree, with no serious failure to replicate.
Supported with caveats: Several studies agree, with stated limits such as student samples or lab settings.
Plausible: One study, or indirect evidence from a nearby question.
Speculative: An idea or argument that hasn't been directly tested.
Marketing claim: Something a seller says about their product, without published evidence.

Evidence: PlausibleIn one study, strangers who viewed 100 Facebook profiles estimated the owners' extraversion, agreeableness and conscientiousness accurately, but not neuroticism.
Why this label
One study, or indirect evidence from a nearby question. Demonstrates: Hall, Pennington and Lueders (2013) used a lens-model design: owners self-reported personality, 53 profile cues were coded, and 35 stranger observers estimated each owner's personality; observers were accurate for three of five traits, and certain profile cues were diagnostic. It shows that a profile can carry some real information about a person, not that a viewer can read a person from it. Marketing framing: a profile 'communicates who you are' so curate it for the right impression (common framing, unattributed). Bears on A4: §43 authentic digital presence and §45 interests item; supports making real interests legible, and limits any claim that viewers can infer much beyond a few traits.
What it does not show
- One study of 100 users with 35 observers; setting and country not given in the abstract; no large replication found.
- Accuracy was judged against self-reported personality, which is itself imperfect.
- Neuroticism was not accurately estimated, and the abstract gives no effect sizes.
- Facebook profiles of the early 2010s; not dating apps or Instagram.
Who it applies to
A pattern across groups of people, not a statement about any one person.
Source
- Hall, J. A., Pennington, N., Lueders, A. (2013). Impression management and formation on Facebook: A lens model approach, New Media & Society.
Evidence: Supported with caveatsStudies of speed dating and early romantic interest found that self-reported traits and stated preferences, measured before people met, did a poor job of predicting who a given person would desire; perceptions of the partner mattered more.
Why this label
Several studies agree, with stated limits such as student samples or lab settings. Demonstrates: Joel, Eastwick and Finkel (2017; two speed-dating studies) found that models using 100+ pre-event measures could predict some general tendencies to desire or be desired, but could not predict desire for specific partners; Eastwick et al. (2022; 208 singles, 7 months) found robust effects of perceptions of the partner and no ideal-preference-matching effect. See attraction-foundations-c3 and evolutionary-mating-preferences-c3 for the stated-versus-revealed meta-analytic picture. Marketing framing: attraction is predictable from a person's type or checklist (common framing, unattributed). Bears on A4: §49 claim audit; supports 'what people say they want is not what they respond to', and limits any lesson that predicts attraction from a profile.
What it does not show
- Both studies come from the same research group; the abstracts describe speed-dating and single-adult samples but not their country or age range; replication by other groups was not checked.
- Machine-learning prediction of self-reported interest; null for relationship variance in these designs does not prove compatibility is unpredictable in principle.
- Self-report outcomes, not later relationship success.
- Variance-explained figures are not weights for how much a trait 'matters' and must not be shown as such.
Who it applies to
A pattern across groups of people, not a statement about any one person.
Sources
- Joel, S., Eastwick, P. W., Finkel, E. J. (2017). Is Romantic Desire Predictable? Machine Learning Applied to Initial Romantic Attraction, Psychological Science.
- Eastwick, P. W., Joel, S., Carswell, K. L., Molden, D. C., Finkel, E. J., Blozis, S. A. (2022). Predicting romantic interest during early relationship development: A preregistered investigation using machine learning, European Journal of Personality.

More practice

The rest of the set, Reading someone's profile without guessing: nine profiles: 9 items, about 18 minutes. Say what the profile shows, choose the message it supports, then say what it tells you about interest. Open the exercise set.