How to Reconnect With Someone You Haven't Spoken to in Years
How to reconnect with someone in 5 steps: why your message is likely more welcome than you expect, what to write first and how to turn one reply into a tie.

A name comes up in a meeting, and you think of the colleague who taught you half of what you know about your job. You last spoke six years ago. You open a message, type “Hi, long time”, and close it again, because it feels too strange after so long.
That hesitation is usually misplaced. Across a 2022 series of preregistered experiments, Peggy Liu of the University of Pittsburgh and her co-authors found that people who reached out to someone they had not been in touch with for a while consistently underestimated how much the other person appreciated it, and the gap was widest with people they were not close to.1 So how to reconnect with someone, in short: send a brief, warm note soon, then give the renewed contact somewhere to go. Five steps:
- Choose someone you parted with on good terms
- Write a short note that says why they came to mind
- Keep the apology for the silence to one line
- Suggest a call if the reply is warm
- Agree on the next contact before you say goodbye
A first note can be three sentences long: “I was at the budget review today and remembered how you rebuilt our forecast in a weekend. I still use your template. How are you, and what are you working on now?” It names a real memory, says why it mattered and ends with a question that is easy to answer.
This guide belongs to the site’s networking section, and it goes deep on one practical question: the first message after a long silence, and what happens next.
Why reaching out feels more awkward than it is
Reaching out after a long silence feels risky because the sender pictures the message landing badly, while the recipient mostly feels pleasantly surprised. Liu and colleagues’ 2022 experiments found that recipients valued a simple check-in more than senders predicted, and that the surprise itself helps explain why.1
The study
Moderate evidence
Thirteen studies of the unexpected check-in, 2022
In one campus field experiment, 54 pairs of students took part: one wrote a short note to check in with a classmate they had not been in touch with for a while and predicted how much it would be appreciated, and the classmate then rated it. On a 7-point scale, senders expected about 5.6; recipients reported about 6.2. The same underestimate appeared when people recalled real reach-outs, sent a note with a small gift, or imagined sending one.1
The effect is modest rather than dramatic. Almost everyone in these studies rated a reach-out as welcome; the error is in how welcome. An independent replication of one of the online studies by the Transparent Replications project at Clearer Thinking, a report that was not peer reviewed, found the same gap among 742 people and described the effect as small.2 Small is still useful here, because the decision it tips is whether you send the message at all. Note, too, what “a while” meant: participants chose whom to picture, so few of these studies tested silences of years; the phone-or-email experiment in step 4 is the one built around gaps of at least a couple of years.
Why do senders get it wrong? Liu and colleagues’ evidence points partly to surprise: the recipient’s attention goes to the surprise of being remembered, and surprise amplifies a pleasant moment, while the sender, who knows the message is coming, thinks less about that.1 A 2022 review by the University of Chicago behavioral scientist Nicholas Epley and four co-authors adds a second reason: people judge their own outreach on competence (did I word this well?) while recipients judge it on warmth (they thought of me).3
The same pattern shows up well beyond old friends. In Epley and Juliana Schroeder’s 2014 field experiments, Chicago-area commuters randomly assigned to talk to a stranger reported a more positive trip, while other commuters predicted the opposite.4 The review argues that such beliefs persist because avoiding people means never collecting the evidence that would correct them.3
So if you catch yourself redrafting a two-line message for the fifth time, send the version you have. The wording you are polishing is the part the other person is least likely to notice.
The people you hesitate over most may welcome you most
The underestimate was largest for acquaintances. In the last experiment in Liu and colleagues’ paper, run online with adult panels in the US and UK, people underestimated how much a close friend would appreciate hearing from them, but they underestimated it roughly twice as much for a weak tieweak ties: Relationships with little shared time, emotion, confiding or exchange of favors, such as a former colleague. Because acquaintances move in other circles, they are more likely than close friends to pass on news you have not already heard, including about jobs.Full entry in the glossary, someone friendly but not a confidant. A small pretest found that a message from an acquaintance feels more surprising, which fits the surprise explanation.1
The reverse case supports the same explanation. In another online scenario experiment, when the reach-out was expected, as in a program that pairs people to check in, the gap disappeared: recipients still said they would appreciate it, but no more than senders predicted.1
- Myth
- After years of silence, a message out of the blue will seem strange or needy.
- Fact
- In Liu and colleagues' experiments, recipients appreciated check-ins more than senders expected, and more so when the message was a surprise.
In practice, the former teammate from two jobs ago, the course-mate you got on well with and the supplier who once went out of their way for you are exactly the people you are most likely to talk yourself out of contacting. They are also the kind of acquaintance that Granovetter’s research on weak ties links to news and openings your inner circle rarely hears about. Old contacts can be practical too: in a 2011 study of Executive MBA classes in the US and Canada, executives who were initially reluctant to reconnect rated advice on a work project from contacts they had not spoken to for at least three years favorably against that of current ones, though they chose whom to ask.5
The lesson is to treat your hesitation as a poor guide. When a name keeps coming back to you, that is a reason to write, not a sign that it is too late.
Start with the one you knew least
Write down three people you liked working or studying with and have not heard from in years. Put a star next to the one you knew least well. One online experiment suggests that is where your guess about their reaction is most likely to be too low, so start there.
How to reconnect with someone, step by step
How to reconnect with someone comes down to a warm first message and a second contact you actually arrange. The research speaks most directly to the first two steps; the rest are practical suggestions built on smaller studies or on none, labelled as such.
1. Choose someone you parted with on good terms
Liu and colleagues asked people to contact someone they usually had pleasant encounters with, and they raise the possibility that after negative past encounters the surprise could be unwelcome, and senders might then overestimate how well the message is received.1 Start with the easy cases: the people you simply drifted away from.
2. Write a short note that says why they came to mind
The reach-outs Liu and colleagues tested met a deliberately low bar: a note to check in and say hello, “just because”.1 You do not need a reason, a favor or news. Naming what brought them to mind makes the note personal and gives the other person something easy to answer.
Gratitude works well if you have it. In an experiment described in the 2022 review, MBA students who wrote thank-you letters underestimated how positive their recipients would feel and overestimated how awkward it would be for them.3
Two more notes you could adapt, alongside the one near the top (illustrations, not tested scripts):
- “Your name came up at a conference in Lisbon and I realized it has been far too long. I would love to hear what you have been up to.”
- “I just read an article on hospital staffing and thought of our late-night arguments about rotas. Are you still in the field?”
3. Keep the apology for the silence to one line
No study has tested apologies for lost contact, so this is a suggestion. The review’s point about warmth and competence applies: the recipient is mostly registering that you thought of them.3 If you both drifted, “It has been far too long” covers it. A paragraph of explanation turns a warm note into a confession the other person has to respond to.
4. Suggest a call if the reply is warm
Text feels safer, but in one experiment a voice reconnected people better. In a 2021 field experiment, Amit Kumar and Nicholas Epley asked University of Chicago students to reconnect with an old friend they had not spoken to in at least a couple of years, and randomly assigned them to phone or email. Most preferred email, and the awkwardness they expected from a call best predicted that choice, yet those who phoned reported a stronger bond and felt no more awkward than those who wrote.6
The sample was small and young, and only about half the students completed the task, but in the same paper, laboratory chats between strangers showed the same advantage for voice over text.6 A first message by text or email is fine; once they reply warmly, offer a short call or a coffee.
5. Agree on the next contact before you say goodbye
This is where the evidence runs out. Liu and colleagues measured appreciation right after a single message, not whether the relationship grew closer, and they list that as an open question.1 They also suggest people sometimes mean to reach out, then put it off and forget.1
- 1First notewhy they came to mind, one easy question
- 2Their replyanswer the same day if you can
- 3A call or coffeevoice over text once it is warm
- 4A dated next steplunch, an article, an introduction
So make the next step concrete: a date for lunch, an article you promised to send, an introduction you offered. A short list of people and the date you last spoke, sometimes called a personal CRM, keeps these promises from slipping. Friendships fade when contact stops, as the research on Dunbar’s number and relationship decay describes, so one good conversation is a restart, not a finished repair.
Your reconnection note, checked
When a message may not land well
Most of the evidence concerns pleasant past relationships and messages that ask for nothing. Three situations sit outside it.
After a conflict, the authors’ own caution applies: they suggest that after negative past encounters, a surprise message may feel unsettling rather than warm.1 If the relationship ended badly, a longer, more careful approach, or no approach, may be wiser.
Asking for something straight away changes the message. That is not wrong: Liu and colleagues count a request for help as a reach-out too, but note that it adds an element beyond simply checking in.1 Say what you want in plain words and keep it small, so the other person can say no easily.
Culture matters too. Most of these studies took place in the US, with some in the UK and Canada. Epley’s review notes that, in one experiment, American and Chinese participants’ expectations about asking people for help differed more than their actual experiences did.3 Adjust the tone and formality to local norms; the underlying point, that people tend to expect a cooler response than they get, has appeared in both the US and UK samples tested so far.
How much of this advice rests on research
Steps 1 and 2 have the firmest support in this guide, a finding repeated across the nine main experiments of one preregistered paper and in one independent replication: people underestimate how welcome their check-in will be. The advice on whom to contact first and on calling rather than emailing rests on single studies, and keeping the tie going rests on no study yet.
| Idea | What the best evidence found | Evidence |
|---|---|---|
| Recipients value a check-in more than senders expect | Underestimate found across recall, campus field and online studies, and in one independent replication | Trial, moderate; small effect12 |
| The underestimate is larger with acquaintances | Larger gap for weak ties than close friends in one online experiment | Trial, limited (one scenario study)1 |
| Senders judge competence, recipients warmth | Pattern across experiments on conversation, gratitude and kindness | Review of experiments, moderate3 |
| A call reconnects better than email | Stronger bond by phone and no more awkwardness, among students | Trial, limited (one small field study)6 |
| One message renews a relationship | Not measured: studies stopped at appreciation of a single contact | Gap1 |
Read the table as a guide to how firmly to hold each step. Send the first note on the strength of the top row, and treat the rest as sensible habits you can bend to the person: if an old colleague would rather email than talk on the phone, nothing in the evidence says the reconnection has failed.
The bottom line
The message you keep not sending is probably more welcome than you think, especially to someone you were not very close to. Send a short, warm note that says why they came to mind, move to a call if they answer, and set the next contact before you hang up. The research shows that the first message is appreciated; whether it becomes a relationship again depends on the second and third.
Frequently asked questions
What if they don't reply?
No study has measured what a non-reply means, so treat it as missing information, not a verdict: messages get buried. One gentle follow-up after a couple of weeks seems reasonable; after that, let it rest. In Liu and colleagues' experiments, almost every participant rated a check-in as appreciated, so do not read rejection into a quiet inbox.
Is it all right to reconnect when I need something, such as job advice?
Yes, if you are honest about it. Liu and colleagues, whose 2022 experiments tested simple check-ins, count a request for help as a reach-out too, though it adds an element of asking for help beyond simply checking in. Say plainly what you are asking, keep the request small, make it easy to decline and show interest in the other person's life as well.
Can I reconnect over LinkedIn or social media?
You can start there, since a platform message is often the only address you have. No study compares LinkedIn with email or text for reconnecting. The closest evidence, a 2021 experiment by Amit Kumar and Nicholas Epley, found that hearing a voice built a stronger sense of connection than typing, so treat the platform message as the door, not the whole conversation.
Does this work across cultures?
The evidence comes mostly from the US, with some studies in the UK and Canada. A 2022 review by Nicholas Epley and colleagues describes one experiment in which American and Chinese participants' expectations about asking for help differed more than their actual experiences did. That fits the same pattern, but it has not been tested for reconnecting. Adjust tone and formality to local norms.
Sources
- The surprise of reaching out: Appreciated more than we think. Liu, P. J., Rim, S., Min, L. & Min, K. E. (2023; online 2022). Journal of Personality and Social Psychology, 124(4), 754-771
- Report #7: Replication of a study from The Surprise of Reaching Out: Appreciated More Than We Think. Transparent Replications (Clearer Thinking) & Metskas, A. (2023). Replication report, not peer reviewed
- Undersociality: miscalibrated social cognition can inhibit social connection. Epley, N., Kardas, M., Zhao, X., Atir, S. & Schroeder, J. (2022). Trends in Cognitive Sciences, 26(5), 406-418
- Mistakenly seeking solitude. Epley, N. & Schroeder, J. (2014). Journal of Experimental Psychology: General, 143(5), 1980-1999
- Dormant Ties: The Value of Reconnecting. Levin, D. Z., Walter, J. & Murnighan, J. K. (2011). Organization Science, 22(4), 923-939
- It's surprisingly nice to hear you: Misunderstanding the impact of communication media can lead to suboptimal choices of how to connect with others. Kumar, A. & Epley, N. (2021). Journal of Experimental Psychology: General, 150(3), 595-607
How we researched this
We read the 2022 Liu, Rim, Min and Min paper, Epley and colleagues' 2022 review, Kumar and Epley's 2021 experiments, Levin and colleagues' 2011 dormant-ties study and a 2023 independent replication in full, and the main parts of Epley and Schroeder's 2014 commuter study, found through Crossref and web searches run in September 2026. The weak spot: nearly all of this work used US students or online panels and measured feelings right after one message, not lasting relationships.


