Everything below is sourced, and the sources are listed at the bottom so you can check them. You're probably going to have to justify this to somebody, and half an argument falls apart the moment they do.
What your team needs to crack it
Every case is built so it can't be solved alone. The evidence is split between people. Nobody is given enough to name the killer on their own. To get there, your team have to do six things.
- Talk to people they don't normally talk to.
- Say what they know out loud, before anyone asks for it.
- Listen to someone junior when that person is right.
- Change their mind when the evidence turns.
- Work out quickly who is good at what.
- Hold the whole picture, not just the bit in front of them.
If your team can't do all six today, that's normal. Most teams can't. They'll have done all six by the time they name the killer.
What happens to teams who stop meeting
When Microsoft moved its US workforce fully remote in 2020, researchers tracked 61,182 employees. The share of collaboration time spent with people outside their own group fell by about a quarter. Networks turned inward. People added fewer new colleagues and stayed with the ones they had.
In the UK, the CIPD found nearly two in five employers worried about inclusion as work went hybrid. When they scored the effects of the shift across everything they measured, organisational culture came out worst.
Nothing visibly breaks when this happens. Teams keep delivering. What thins out is the set of connections that carry information between them, and that's difficult to notice until you need one.
Worth being straight about the limits of that study. It measures communication patterns rather than output, it covers one company in one unusual year, and every author works there. It shows the shape of a network changing. It doesn't show productivity falling.
What makes it work when you put those people back together
There's a seventy-year-old answer to this and most people have never heard of it.
In 1954 the psychologist Gordon Allport set out the conditions under which bringing separated groups together improves how they get on. Fifty years later, two researchers tested the idea across 515 studies covering more than 250,000 people. Contact works. It held across every type of group they examined.
It works best when four conditions are met.
Equal status
Nobody's rank counts for more inside the situation than anybody else's. A job title won't spot the hole in a suspect's alibi. Your newest starter and your MD are looking at the same contradiction with the same tools.
A common goal
Everyone is after the same outcome. Name the killer. One answer, one team, and everybody either gets there or doesn't.
Cooperation rather than competition
The goal can't be reached alone. The evidence is split. No individual is given enough to solve it by themselves, so the only route through is telling each other what they're holding.
Backing from the organisation
The contact is visibly sanctioned rather than improvised. You booked it. Your people are there because the company put them there, which is the condition most social events fail.
We didn't build the format from Allport's list. It came out this way because those are the conditions that make an investigation work as a game. The overlap is exact, and it's the reason this does something a quiz night doesn't.
Why the evidence has to be split up
Groups have a documented habit of discussing what everyone already knows while the decisive detail sits in one person's head. It's been studied since 1985 and it has a name: the hidden profile. Teams get the answer wrong when between them they held everything needed to get it right.
A later review pooled 72 studies covering 4,795 groups and over 17,000 people. Information sharing predicted team performance across every condition they tested. Sharing information nobody else had mattered more than sharing what the room already knew.
Which brings us to the part we'd frame and put on the wall.
In 1992, two psychologists needed a task to test whether groups pool what they know. They gave university students a murder mystery, split the critical clues between them, and had them discuss the case.
Half the groups were told they had enough evidence to identify the killer. The other half were told they didn't, and were only forming a judgement.
The groups who believed there was an answer found the killer 67% of the time. The groups who didn't managed 35%.
Same evidence, split the same way, same kind of people. The difference was believing the thing could be cracked, and it nearly doubled the success rate.
So that's how we build them. Split evidence, a real and findable solution, and a team told from the first minute that it can be solved.
Does this kind of event work?
The largest review of team building interventions pooled 60 correlations and found a moderate positive effect across every outcome it measured. Then it did something more useful. It split interventions into four types to see which were doing the work.
Goal setting scored highest. Role clarification came second. The two that trailed were interpersonal relations and problem solving, which is where most away-day exercises live.
A case is a shared goal with roles that emerge under pressure. It's built from the two components with the strongest evidence behind them, and not from the two without.
Two more findings from the same review. The strongest effects of all were on how teams felt about working together, which is the thing you're being asked to improve. And larger groups benefited more than small ones, which is why we run 30 rather than eight.
Whether the quiet ones speak
Amy Edmondson went looking for a link between good teamwork and fewer mistakes in hospitals and found the opposite. The better teams reported more errors. They weren't making more of them. People felt able to say so.
What separated those teams was a shared belief that you can ask a question, admit a gap or challenge something without it costing you. On the strongest teams that held from the most junior person to the most senior.
A later review pooled 136 samples covering over 22,000 people and around 5,000 groups, and found that belief connected to how well people performed their work.
You already know whether your team has it. What's harder to find is three hours where you can watch it happen.
Whether enjoying it matters
Economists at the University of Warwick ran experiments with over 700 people. They made one group happier, using a comedy clip or an offer of chocolate, fruit and drinks, then set everyone the same paid task.
The happier group produced about 12% more, at the same quality.
It's a laboratory study on an individual task, so we won't pretend it proves a murder mystery moves your quarterly numbers. What it settles is that enjoyment and output are causally connected rather than merely associated, which is the bit people wave away.
What happens on the day
Nobody performs
The suspects do the lying. Your team does the questioning. Nobody is asked to act, read from a script or put on a silly hat. Ever. It's the thing we get asked about most and the answer never changes.
Everyone starts level
Job titles are no use here. Seniority won't help anyone spot the hole in an alibi. Half the time it's the quietest person in the room who cracks it, and everybody sees it happen.
They don't notice they're doing it
No exercises. No flip chart at the end. They share what they've found, argue about what it means, and change their minds because the case demands it. Ask them afterwards what they did and they'll tell you they solved a murder.
It isn't a Zoom quiz
Suspects to interrogate. Evidence landing while they're still arguing about the last piece. Teams breaking off to compare notes and accusing each other in the chat. People forget they're on a video call within ten minutes.
How we compare
Against escape rooms, quizzes and cooking classes
Fun for an hour, forgotten by Friday. Nobody is still talking about the quiz three weeks later. Ours stick because the challenge is real and your team had to work together to get through it.
Against other murder mysteries
Most murder mystery companies ask your team to dress up and act out characters. That's what makes half the room want to call in sick. With us nobody performs. The suspects are handled for you and your team gets on with investigating.
The practical bits
Group size
Every case runs for up to 30 people. That cap is deliberate. Thirty is the number where everyone stays in it, gets their hands on the evidence, and has a real say in who gets accused. Bigger teams split across separate sessions, same day or different days, so 60 people becomes two proper investigations rather than one crowd where half the room drifts off.
What it costs
- A cold case, ready to run: £700 for up to 30 people.
- A case written from scratch for your team: £1,700 for up to 30 people.
- Each additional session of up to 30: £550.
- With professional actors playing the suspects live on camera instead of our AI: from £1,500.
What's included
Our investigation platform, a full digital evidence pack, and a facilitator running the whole thing start to finish. Up to three hours from briefing to reveal. You send your team a link. We handle the rest.
Payment and VAT
50% deposit on booking, the balance the day before.
We don't charge VAT. We invoice from Estonia, so UK business customers account for it themselves under the reverse charge.
Try it before you book
Dead Air is a free ten-minute investigation you can play right now. A podcast host is dead, there are three suspects, and they're the same AI your team would be questioning. Interrogate one and see whether you can trust a word it says.
Sources
- Allport, G. W. (1954). The Nature of Prejudice. Addison-Wesley.
- Pettigrew, T. F. and Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90(5), 751 to 783.
- Klein, C., DiazGranados, D., Salas, E., Le, H., Burke, C. S., Lyons, R. and Goodwin, G. F. (2009). Does team building work? Small Group Research, 40(2), 181 to 222.
- Stasser, G. and Titus, W. (1985). Pooling of unshared information in group decision making: biased information sampling during discussion. Journal of Personality and Social Psychology, 48(6), 1467 to 1478.
- Stasser, G. and Stewart, D. (1992). Discovery of hidden profiles by decision-making groups: solving a problem versus making a judgment. Journal of Personality and Social Psychology, 63(3), 426 to 434.
- Mesmer-Magnus, J. R. and DeChurch, L. A. (2009). Information sharing and team performance: a meta-analysis. Journal of Applied Psychology, 94(2), 535 to 546.
- Edmondson, A. C. (1996). Learning from mistakes is easier said than done. Journal of Applied Behavioral Science, 32(1), 5 to 28.
- Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350 to 383.
- Frazier, M. L., Fainshmidt, S., Klinger, R. L., Pezeshkan, A. and Vracheva, V. (2017). Psychological safety: a meta-analytic review and extension. Personnel Psychology, 70(1), 113 to 165.
- Oswald, A. J., Proto, E. and Sgroi, D. (2015). Happiness and productivity. Journal of Labor Economics, 33(4), 789 to 822.
- Yang, L. et al. (2022). The effects of remote work on collaboration among information workers. Nature Human Behaviour, 6(1), 43 to 54.
- CIPD (2023 and 2025). Flexible and hybrid working practices. Employer surveys, samples of about 2,049 and 2,050.