Key Takeaways
- The team-culture stuff that makes a real difference rarely looks impressive on a slide.
- A weekly custom question β one that mentions actual things happening on your team β beats a thousand-dollar offsite for trust-building. Boringly often.
- The trick is specificity. The question has to feel like it was written for your team, or people answer on autopilot.
- For remote and async teams especially, this might be the one ritual that survives the loss of hallway chat intact.
- Cost: roughly five minutes a week. Reward: the team actually starts to know each other.
Why $50,000 offsites don't fix team culture
Your team's monthly all-hands ended ten minutes ago. You sat through the deck. You watched the engagement numbers tick down again. And in the Slack channel that's supposed to be "for casual chat" β you know the one β the last message is from three weeks ago. It was you. You shared a meme. It got two reactions, one of them yours.
Your team is fine. Nobody's leaving. Nobody's openly unhappy. But nobody's connected either. And you can feel it.
You're not imagining the gap. And β for what it's worth β most of the obvious solutions don't close it. A founder I know spent fifty thousand dollars on a leadership offsite last year. Two chefs, a thoughtfully-curated wine list, four time zones reunited in a mountain lodge. Their team came home glowing. Six weeks later, the same team felt exactly as disconnected as it did before the planes took off.
This is not a story about offsites being bad. Offsites are great. They produce real, vivid moments. But they're like fireworks β beautiful, brief, gone. They don't survive the gravitational pull of a normal workweek.
What does survive it is something else entirely. Something so unimpressive-looking that most leaders never even try it. It costs basically nothing. It looks like a meeting opener. And the reason it works is almost embarrassingly simple β it's because it looks like it shouldn't work.
It's a custom question, asked once a week, with intention. And the gap between a generic one and a custom one is way bigger than anyone gives it credit for.
Why custom questions outperform generic icebreakers
Imagine two different Mondays at your team's stand-up.
Monday A: "What's everyone's favourite hobby outside of work?" You can already hear it. Hiking. Reading. "Spending time with family." Nothing wrong with any of that β but nobody learned anything new. The question got answered. The team didn't move.
Monday B: "If we kept doing ONE weird thing about how this team works, forever, what would it be?" Now the room gets interesting. Someone says "the way Maya always opens her camera before her video actually kicks in." Someone says "the fact we keep accidentally inventing the same Star Wars references in code reviews." Someone says "the silence after Dan asks a hard question β that pause is sacred, don't ruin it." You learn that your teammates have noticed each other. That changes things.
That's all a custom question really does. It says, in code, "I see this team. I see you." When that signal is convincing β when the question references real moments, real running jokes, real projects β people respond with more of themselves. Harvard's Alison Wood Brooks and Leslie John, writing in HBR, make this point more rigorously: asking good questions is one of the most underrated tools for building rapport, and most professionals dramatically underestimate it as a skill.
The rest of us call it "the weird Slack thread where I learned more about my colleagues than in a year of weekly one-on-ones."
Why custom questions matter most for remote teams
In an office, this stuff happens by accident. Someone mentions their dog ate the leftovers. Someone overhears it. Someone else asks about the dog two days later. Multiply by a thousand small exchanges and a team comes to know each other almost incidentally. Remote teams don't get that for free.
Slack is great, but it has a "should I even bother posting this?" tax that doesn't exist in a kitchen. The little things stop happening, and after a while the team starts to feel less like teammates and more like people who happen to share a Notion workspace. GitLab β about as async-first as a company gets β has spent years documenting how remote teams replace those moments deliberately, in their public handbook. It's worth a read even if you never work there.
I've written before about the difference between coordination and cohesion in distributed teams. The short version: you can have a team that ships beautifully and still feels like a collection of strangers. The fix isn't more meetings. The fix is restoring the small, repeated, slightly-personal moments that used to happen by accident.
A weekly custom question is a hallway moment, scheduled. Not as good as the real thing β nothing is β but durable, deliberate, and an order of magnitude better than the absence of it.
How to actually do it (it's almost embarrassingly simple)
- Pick a cadence and protect it. Once a week is the sweet spot. Often enough that the ritual builds momentum, not so often that it becomes another notification to dismiss.
- Write questions that reference your team. This is the whole game. Name the project. Reference the in-joke. Mention the moment from the all-hands. The question should make people smile a little before they even answer it. (Examples in a second. If writing one every week sounds like friction you won't sustain, Icebreakerz automates exactly that β but more on that below.)
- Rotate who writes them. If one person writes every week, the questions will quietly start reflecting that person's interests. Rotate. The teammate writing this week's question gets a small dose of creative agency β itself a culture-building act.
- Keep the bar low. A custom question is not a performance review. One sentence is fine. A gif is fine. Three words and a shrug emoji is fine. The point is the ritual, not the depth.
- Read every answer. Especially leaders. If a teammate writes something and nobody seems to have read it, the ritual decays fast. One emoji reply is enough. Silence is fatal.
Examples of custom questions that actually land
- "Who said 'I'll fix it Monday' four Mondays in a row about the same bug β and what finally happened?"
- "What's the most unhinged feature branch name anyone committed this year, and did it actually ship?"
- "Which Slack thread from Q3 started as a 2-minute question and somehow turned into a 47-message philosophical debate about naming conventions?"
- "Which of our running jokes would a new hire need to be onboarded into within their first week or they'd be completely lost?"
- "Name a feature you've quietly wanted to suggest but haven't because you weren't sure if it was a 'real' idea yet."
- "Who on this team is most likely to start an unprompted Slack thread about something incredible they just read? What's the topic going to be?"
- "If we had to pick one weekly meeting to delete forever β and replace it with anything we wanted β what would we replace it with?"
Notice what these have in common: they make your teammate feel like the team has a history. That's the trick. Even when the team is new, asking specifically creates the texture.
What your team feels like after three months of this
The shared channel that used to be tumbleweeds-and-the-occasional-meme has actual threads in it. People reference each other's weekend hobbies in standups casually, because they actually know them. There's a running joke that someone planted in week four, and the team carried it forward without anyone organizing it. When someone joins from a different team, the standup feels weirdly intimate to them β there's texture you can feel in five minutes.
Three months. From "fine but a little disconnected" to "this team has its own gravity." All from five minutes of writing-a-question per week.
And here's where the contrast gets sharp: the team that did this and the team that spent $50K on the offsite are both trying to solve the same problem. One of them moved. One of them didn't. The cheap one moved.
So what makes them the sneaky-best part of team culture?
Most of what works in team culture is also obvious. Hire well. Run good 1:1s. Have a clear mission. Compensate fairly. None of that is news.
The sneaky part of custom questions is that they look like the opposite of culture work. They look like the throwaway icebreaker a manager opens a meeting with because they once read a Medium post about engagement. They look like nothing. So most leaders dismiss them. So most teams never invest in doing them well.
But the actual mechanism underneath β small, repeated, low-stakes disclosure with specificity that signals "I see this team" β is doing the same job a five-figure offsite is trying to do. Just at one-thousandth the cost and with thirty times the repetition over a year. The offsite is a fireworks show; the weekly custom question is the long, slow, almost-imperceptible drip that actually changes how teammates feel about each other six months from now.
That's the trick. It's mistaken for fluff. Which means the teams that do run it β quietly, consistently, with actually-specific questions β get a real, compounding culture advantage everyone else assumed wasn't worth the time. It's hiding in plain sight, dressed as fluff, doing real work. That's what makes it the sneaky-best part of team culture.
One small note on tools
You don't strictly need a tool for this β a shared doc, a Slack channel, a recurring calendar invite all work. But the friction of writing a good custom question every week is what quietly kills rituals, which is the gap Icebreakerz was built to close β more on that below. The shape of the tool matters less than the commitment to the practice. If you take one thing from this post: pick a cadence, write your first custom question for your team this week, and run it.
FAQ
What counts as a "custom" question?
A question that references something specific to your team β a project, a person, a recent moment, an inside joke. Anything that signals it was written for this team and not lifted from a list of 100 icebreakers.
How often should we run them?
Once a week is the right starting cadence. Less loses momentum; more competes with actual work for attention.
Won't this feel forced at first?
Yes. The first two or three will be awkward. Push through. The awkwardness fades fast once people see colleagues answering honestly.
What if nobody engages?
Two reasons: the questions are too generic, or leaders aren't visibly participating. Both are fixable. Tighten the questions, show up, and engagement follows.
Can AI write the questions for us?
Sort of. An AI prompted with "give me a team-building question" produces something forgettable. An AI given real context β projects, dynamics, in-jokes β can produce surprisingly good ones. Treat it as a junior writer that needs a good brief, not a magic button.
Sources
- The Surprising Power of Questions β Alison Wood Brooks & Leslie K. John, Harvard Business Review, MayβJune 2018. Research on how asking questions builds rapport, trust, and openness. (Cited in "Why custom questions outperform generic icebreakers".)
- GitLab Handbook β Team Building β GitLab Inc., 2025. A working playbook from a fully-async company on building culture through structured rituals. (Cited in "Why custom questions matter most for remote teams".)
About the author
Luke T. is a senior software engineer and founder of GoFish Gallery, based in the US. After years of remote work and sitting through countless virtual meetings that felt disconnected and transactional, he started building tools to fix what most companies just accept. Stellar Bonds came from a simple frustration: remote teams deserve more than another Zoom call. He builds games and rituals that make distributed teams actually feel like teams β Stellar Bonds, Icebreakerz, and more.
Want to actually run this with your team?
Icebreakerz lets you set up custom recurring questions in a few minutes β no signup wall, no per-seat pricing for small teams.