When Your Adult Has No Friends: Practical Steps for Connection
Friendship is not a count. Connection is what the post is about. Why the social-skills-training era over-counted, what supported social opportunities actually deliver, and what to measure instead of friend totals.
1
One real connection outweighs ten polite acquaintances. The goal is the relationship the adult chooses to maintain.
Friendships form around shared interests. Coaching the interest is coaching the friendship. Not what we wish they cared about.
Hobby groups, faith communities, gaming circles. Lower social tax. Higher signal-to-noise. Mistakes are normal.
Drop off. Step back. Pick up. Repeat. The caregiver’s job is transportation, not facilitation.
List three genuine interests tonight.
Real interests. Not what we wish they cared about. Not what looks normalizing on paper. What does the adult actually pull toward when nobody is supervising. Trains. Specific anime. Cooking shows. A particular musician. Birds. Magic the Gathering. Whatever the answer is. The list is the brief. Friendships form around shared interests, never around general “social skills.”
Friend count is the wrong metric
Counting “friends” produces miserable adults with twelve polite acquaintances and zero real connections. Pick a different metric. Count chosen returns, not headcount.
Interest precedes social skill
The shared interest is the magnet. Social skills are the surface. Coach the interest and the surface usually follows. Coach the surface alone and the relationships rarely form.
Forgiving environments
Spaces where mistakes are normal, expectations are clear, and conversation centers on the shared interest, not the people. Hobby groups outperform “social skills groups” almost every time.
Connection takes time
Real connections form over six to twelve months of recurring contact. The first three months can look like nothing is happening. Stay through them.
Caregiver as logistics
Drop off, step back, pick up. Stay nearby for the first few visits, then fade out. Resist real-time facilitation. Connections form when the caregiver is not in the room.
Measure what matters
Did the adult initiate contact this week? Did they choose to return? Did they describe the interaction positively? Those are real signals. “How many friends do you have” is not.
Where connection forms
He has one friend. His name is Trevor. They met at the model train club at the library three years ago. They talk about Lionel trains for forty-five minutes every other Tuesday. They never see each other socially outside of the club. Trevor has never been to our house. My son has never been to Trevor’s. They are friends. The fact that there is only one of him is not a problem. The problem was assuming we needed more.
Building connection over six months
- Identify three specific interests the adult genuinely cares about (not aspirational).
- Find one recurring group for each. Visit each once over three weeks.
- Choose the most workable group. Commit to attending for three months.
- Drop off, step back, pick up. Resist real-time facilitation.
- At month three, check the signal: initiation, return, positive description.
- Interest inventory (real, not aspirational)
- Three candidate groups identified
- First visits scheduled
- Six-month commitment to one group
- Caregiver-as-logistics rule in place
- Monthly signal tracking (initiation, return, description)
- Online safety routine (if applicable)
Friendship is one of the harder things to engineer and one of the easier things to obstruct. Engineer it carefully. Obstruct it never.
The full story · For readers who want context
Frank’s son has one friend. His name is Trevor. They met at the model train club at the Escanaba public library three years ago. They talk about Lionel trains for forty-five minutes every other Tuesday in a meeting room next to the children’s section. They never see each other socially outside of the club. Trevor has never been to Frank’s house. Frank’s son has never been to Trevor’s. They exchange a Christmas card every year. They are friends. Frank knows this because his son talks about Trevor between Tuesdays, asks if Trevor will be at the next meeting, and seems quietly delighted when the answer is yes. The fact that there is only one of him is not a problem. The problem, the one that took Frank and his wife years to unlearn, was assuming they needed more.
Friend count is the wrong metric.
Here is what they will not tell you on the front of any social-skills curriculum. The dominant model for “improving social outcomes” in autistic adults for the past several decades has been to coach the surface behaviors of friendship (eye contact, conversational turn-taking, small talk, reciprocity) on the implicit assumption that mastering these will produce friends in the way mastering arithmetic produces correct sums. The model is wrong. The model is wrong in a specific way: it confuses the surface of friendship with the substance of it.
The substance of adult friendship, autistic or otherwise, is recurring contact organized around shared interest. The surface is whatever the dominant social culture says it should look like. Coaching the surface without addressing the substance produces adults who can perform “friendliness” without forming friendships, and who then describe themselves as lonely despite having checked every box that a social-skills curriculum says they should have checked. Friend counts go up. Self-reported loneliness stays the same or worsens. This pattern has been documented in the autism research literature, particularly in the work coming out of the AASPIRE collaboration and Drexel’s A.J. Drexel Autism Institute on autistic adult social outcomes.
Interest precedes social skill.
Adult friendships in the general population form, almost always, around shared activities. Coworkers who eat lunch together. Members of the same hobby club. Regulars at the same coffee shop. Neighbors who walk dogs at the same time. The activity is the prior. The friendship is the consequence. Strip out the activity and the friendship rarely forms; provide the activity and the friendship often does, even between very different people, because the activity creates the recurring contact and provides the topic.
Autistic adults are no different. The shared interest is the magnet. A model train enthusiast in his fifties and an autistic adult in his late twenties can have a forty-five-minute conversation about Lionel engines that neither of them would be capable of having about the weather. The interest does the work that small talk does for neurotypical adults. Frank’s son and Trevor have a forty-five-minute conversation about trains. They do not have a forty-five-minute conversation about themselves. Both of them would describe the conversation as good.
Forgiving environments.
Most “social skills groups” run by clinicians are simulations of social environments. They are not the social environments themselves. The simulation has limited transfer to real settings because the real settings have different rules, different conversation topics, and different stakes. Hobby groups, faith communities, gaming circles, and volunteer corps are real settings. They are also forgiving in ways that bars, parties, or generic social events are not. The reasons: the conversation is supposed to be about the activity, not about the people; mistakes in conversation are absorbed by the activity moving forward; expectations are clear and bounded; the social cost of being awkward is low because nobody is there to evaluate the people.
Choose forgiving environments first. Bridge to less forgiving ones later if the adult wants to. Most adults do not need to. The forgiving environment, attended regularly for years, produces durable connections that the adult experiences as friendship even if a neurotypical observer would not call them that.
The caregiver’s job is logistics.
This is the part that most caregivers find hardest. The caregiver’s role in adult social development is logistics: identifying the candidate environment, arranging transportation, providing financial coverage, supporting recovery from setbacks. The caregiver’s role is not facilitation. The caregiver should not be in the room translating, redirecting, repairing conversation, or prompting interactions. Connections form when the caregiver is not visibly present.
Stay nearby for the first few visits if needed. Fade out over four to six weeks. Drop off at the door, sit in the car or in a coffee shop for the hour, pick up at the end. Resist the urge to coach in real time. Resist the urge to ask “did you make any friends today” on the drive home. Ask, instead, “what did you talk about.” If the answer is about the activity, the work is happening. If the answer is “nothing,” try again next week. The connections form on their schedule, not yours.
Six months, not six weeks.
Real connections form over six to twelve months of recurring contact. The first three months often look like nothing is happening. The adult attends. The adult sits next to the same person. The adult exchanges a few words. The adult goes home. The pattern can look like failure for months. Then, somewhere around month four or five, the adult mentions the person’s name without prompting. Then the adult notices when the person is absent. Then the adult is quietly delighted when the person is present. That is the connection forming, on its own clock, without anyone having coached it into existence.
Stay through the first three months. Most families abandon at week five or six because nothing visible is happening. That is the worst possible time to abandon. The connection that would have formed at month five is what the abandonment prevents.
Online connection: real, with safety.
A meaningful share of autistic adult friendship now forms online. Discord servers organized around specific games, fan communities for specific shows, forums about specific interests, online tabletop play. These connections are real. They count. They are also not without safety considerations: financial exploitation, identity theft, romance scams, predatory grooming, and other harms target socially isolated autistic adults online at elevated rates. We covered the financial scam pattern specifically in another post; the same caution applies to online social settings.
Set up the safety routine before the online community attendance. Monitored email account or shared family Discord login for younger adults still building independence. No financial information sharing. No address sharing. No in-person meet-ups without family verification of the other person. Periodic check-ins by a trusted family member or DSP about what is happening in the community. The safety routine does not preclude the connection; it makes the connection possible without the corresponding risks.
Measure what matters.
The right metrics for adult social connection. Did the adult initiate contact with anyone this week (a question, a comment, a return to the group). Did the adult choose to return to a setting they have visited before. Did the adult describe the interaction positively when asked. Does the adult identify with the group (“I’m in the model train club”). Is there at least one named person the adult talks about between visits. Does the adult notice when that person is absent.
These six signals are more diagnostic than any friend count, party invitation total, or social skill assessment score. They track what is actually happening in the adult’s social life rather than what looks normative on a clinical metric. The adult with one named friend whose absence is noticed has a richer social life than the adult with twelve acquaintances whose names blur together.
What other states make easier.
Several state Arc affiliates and developmental disabilities councils fund inclusive social programs explicitly modeled on shared-interest rather than social-skills curricula (Massachusetts, Pennsylvania, Ohio). The Best Buddies program operates state chapters that match adults with disabilities to neurotypical “buddy” volunteers for one-to-one friendship over months (national, with strong state presence in California, Florida, Texas). Special Olympics Unified programs build friendships around shared sport participation. Online, communities like Wrong Planet have hosted autistic-adult-led social spaces for years. Michigan funds Community Living Supports that can fund DSP hours to accompany adults to community settings, but with less coordinated state-level inclusive social infrastructure than peer states. Other states made different choices. Michigan didn’t, fully.
Your assignment this week.
Tonight: the three-interest list. Real interests, not aspirational. This week: identify one recurring group for each. Within two weeks: visit each once. Within four weeks: choose the most workable and commit to six months. Within six months: check the six signals. None of this is exotic. None of it was supposed to be hard to learn. Other states make the infrastructure more accessible. Michigan didn’t, fully. Now we know what we are fighting. Together.