How to Run a Company Game Show in Singapore

Building a Company Quiz: Rounds, Buzzers and the Wager

In Singapore the corporate game show is sold by at least five distinct suppliers, split between dedicated quizmasters who host pub quizzes on weeknights and take corporate bookings on the side, and full-service team building operators who package the same thing as a virtual or in-person product. Published rates for the packaged versions run from about $20 to $45 per person online and $25 to $50 per person for an in-person indoor session, while several of the specialist quizmasters publish no rate at all and quote on enquiry. Sessions are typically 60 to 90 minutes, teams of four to six, and one local supplier publishes a standard six-round structure with a wagering round in it.

First, the disclosure that belongs at the top: PLAYON does not run this format, and there is no facility here for it. That is the whole of the connection, and the rest of this page is written for whoever has to build or buy the quiz.

Everything published about this format describes the wrapper. Vendors sell you a host with a microphone, a slide deck, a music round and a leaderboard. What none of them publish, and what actually determines whether your afternoon works, is the question set and the round structure. A quiz is nothing but questions in an order with points attached. Get the questions wrong and you have run a two-hour public demonstration that two people in your office read a lot. Get them right and you have thirty people arguing productively in five groups. This page is about that difference.

What a corporate game show actually is once the theming is stripped away

A corporate game show is a scored, host-led contest in which teams answer sequenced questions or complete short televised-style challenges, with points accumulated across rounds and a single winner declared at the end.

The word carrying the weight is sequenced. A quiz is not a pile of questions. It is an ordered sequence in which each round does a specific job: settle the room, level the field, create a shock, restore hope, decide the thing. That ordering is the craft, and it is invisible in every listing you will read.

Local supply divides into three shapes, and the shape tells you what you are buying.

The specialist quizmaster. An individual with a long history of running public pub quizzes who also takes corporate bookings. They bring the questions, the answer sheets, the scoring and themselves. You bring the room, a PA system and the prizes. Published capacity from local specialists runs to 30 or 40 teams, with a stated floor of about 20 participants and a comfortable ceiling above 200. Run times published locally cluster at 90 minutes. Rates are almost never published and are quoted per event rather than per head, which is the correct pricing shape for this format and worth knowing before you compare it against per-pax quotes.

The packaged product from a team building operator. The quiz is one item in a catalogue of thirty. It arrives with branded slides, a facilitator rather than a quizmaster, and a per-pax price. Published local figures for online editions include entries from $20 per person for 1.5 hours with no stated minimum, from $25 per person for one hour at 10 to 80 participants, and from $40 per person for 1.5 hours at 16 to 60 participants. In-person indoor sessions of this intensity publish at $25 to $50 per person for 30 to 90 minutes.

The physical-challenge game show. Less a quiz than a set of televised-format stunts: survey-answer rounds, category boards, price-guessing segments, elimination gauntlets. These are sold under the game show heading and they behave nothing like a quiz. They are closer to a parallel-station event with a microphone over the top. If your brief says quiz and the quote describes props and a stage, you are being sold a different format.

The rest of this page addresses the quiz proper, because that is what most companies actually want when they say game show and it is the version an in-house organiser can genuinely build.

Why general knowledge is the wrong spine for a workplace quiz

Here is the central flaw of the format, stated plainly. A general knowledge quiz has a known winner before it starts.

Every office contains two or three people who consume a lot of information and retain it. In a pure general knowledge quiz, those people answer, their teams win, and the other twenty-five people spend ninety minutes as witnesses. Worse, they learn something they did not want to learn, which is where they rank on a public leaderboard of recall. That is the opposite of what the afternoon was booked to achieve.

The structural reason is that general knowledge is a lookup task. Either the fact is in your head or it is not. There is nothing for a team to do with a lookup question except poll its members and take the one confident answer. No conferring happens, because there is nothing to confer about. The team is a delivery mechanism for its best-informed individual.

Three families of question break that pattern, and each breaks it differently. They are not interchangeable and a good round order uses all three.

Company-specific questions move the knowledge base from the world to the organisation. Suddenly the person who has been in the finance team for eleven years knows things the well-read newcomer cannot. They redistribute expertise rather than removing it, which is why they are the single highest-value category and also the most work to write.

Identity and recognition questions, of which guess-the-baby-photo is the archetype, remove knowledge entirely and replace it with attention. Nobody has studied for these. They produce noise, laughter and cross-table shouting, and they cost the participants nothing in status when they get one wrong. They are the social lubricant of the format.

Estimation questions, where the answer is a number nobody knows exactly, change who talks. They are the most underused device in corporate quizzing and they get their own section below.

To be fair to general knowledge: it is not banned, it is demoted. One round of it, placed where it can do least damage, is useful. Six rounds of it is the format failing.

Question categories that suit a mixed office, and the ones that quietly fail

Sort every question you write into one of these. If your set is lopsided, fix it before you worry about difficulty.

Categories that work

Company facts, verifiable and neutral. Which year did we open the second office. How many people are in the company today, to the nearest ten. What is written on the poster by the pantry. Which product line launched first. Rules for writing these: the answer must be checkable in a document you can display, it must not be commercially sensitive, and it must not be knowable only by the leadership team, or you have built a general knowledge round with a smaller pool of experts.

Colleague identity. Baby photos, childhood photos, desk close-ups, handwriting samples, first jobs, home countries, a recorded voice saying one sentence. Consent is not optional here. Ask every person whose photo or voice you plan to use, individually, in advance, and accept a no without negotiation. Anyone who declines simply is not in the round.

Estimation. Any question whose answer is a specific number that almost nobody will know precisely. Covered in full below.

Sensory and recognition rounds. Music intros, film stills, logos with the wordmark removed, close-cropped images, sound effects, anagrams of colleagues' names. These reward pattern recognition rather than study, and the pool of people who can contribute is much wider than for text trivia.

Ordering and ranking. Put these five events in chronological order. Rank these five countries by population. Ordering questions are the quiet star of the genre, because partial knowledge is genuinely useful: a team that knows two of the five positions can reason out a third. That is conferring made mechanical.

Puzzle and lateral questions. Wordplay, connections, spot-the-link, a picture that resolves into a phrase. Nobody has revised for these and they can be solved from a standing start inside a round.

Categories that fail

Broad general knowledge as the backbone. For the reasons above.

Specialist categories chosen by the organiser's own taste. A round on Premier League football or on nineties Cantopop is a round designed to be won by the people who already share the organiser's interests. If you must run a themed round, run two short themed rounds with deliberately opposed themes so that no single subculture takes the whole segment.

Anything culture-locked in a country where the office is not. Singapore workplaces routinely contain people who arrived two years ago and people born here. A round on local television from the 1990s is a round with a built-in scoreboard by tenure and nationality. The same goes for questions that assume one language, one school system or one childhood.

Questions where the answer is a long string. Full names of treaties, exact chemical names, seventeen-word film titles. They mark badly, they generate disputes over spelling and they reward nothing except handwriting.

Anything with a contested answer. Rankings that changed last month, records that depend on the source, superlatives with more than one defensible holder. These are dealt with in the one-answer section below.

Personal questions that can embarrass. Marital history, salary, health, anything that a colleague would not volunteer at a table of strangers. The line is simple: if you would not print it on a name badge, it does not go in the quiz.

The estimation round, and what it does to who speaks

This deserves proper treatment because it is the one round type that structurally forces a team to behave like a team.

An estimation question has a precise numerical answer that essentially nobody in the room knows: the height of a named building in metres, the number of coffee capsules the office got through last year, the weight of an adult manatee, the number of steps from the lift lobby to the far meeting room, the year a company was founded. Scoring is by proximity, not by correctness.

Three things happen that do not happen in any other round.

Nobody is disqualified from having an opinion. In a lookup question, a person who does not know the fact has nothing to say and correctly says nothing. In an estimation question, everyone has a guess, because everyone can reason from something. That single property converts silent participants into contributing ones, and it is why the estimation round is the closest thing this format has to a genuine inclusion mechanism.

The team has to aggregate rather than defer. With five guesses of 40, 60, 120, 150 and 400, the team must decide what to write down. It cannot simply take the loudest number. The obvious method, averaging, is often the right one, and a group that discovers this in round four will apply it in round six. Teams that instead hand the pen to the most confident person consistently underperform teams that average, which is a lesson you can point at in the debrief with actual evidence from your own scoresheet.

Confidence gets separated from accuracy in public. The person who said 400 with total certainty is visibly wrong within ninety seconds, and the person who mumbled 60 is visibly close. This happens without anyone having to say anything about it, which is the only way that observation ever lands well at work.

Writing estimation questions

The answer must be a single unambiguous number with a stated unit and a checkable source you can show on the screen. Aim for questions where plausible guesses span at least one order of magnitude, because a question everyone estimates within five per cent discriminates nothing. Mix scales deliberately: one tiny number, one number in the hundreds, one in the millions, one date, one duration. Include at least two questions whose answers relate to your own building or company, because those are the ones where reasoning from evidence beats worldliness.

Six questions is the right length. Below four the round does not have time to teach anything, and above eight the arithmetic of scoring it slows the event down.

Scoring estimation

Do not score by absolute distance from the answer, because a question with a large answer would then swamp a question with a small one, exactly as an unnormalised raw score does in any multi-unit contest.

Score by rank instead. For each question, order the teams by how close their answer is, then award on a fixed ladder. With five teams: 3 points to the closest, 2 to the second closest, 1 to the third, 0 to the rest. This is fast to do live, it is unit-independent, and it means a wild answer costs a team one question rather than the round.

Two refinements worth having. First, exact answers get a bonus point, which makes the round feel winnable rather than merely survivable. Second, announce ties as ties and award both teams the higher value, because arguing about which of two answers is fractionally closer wastes more time than the point is worth.

Buzzers or answer sheets: the choice that decides everything else

This is the most consequential decision in the whole design, and most organisers make it by accident, based on whether the supplier owns buzzers.

Buzzers reward speed and confidence. In a buzzer format, the first hand down claims the question, and the person who buzzes is the person who both knows the answer and believes they know it. Those are different traits and the second one is not evenly distributed. Buzzers produce a fast, loud, televisual event where four or five individuals across the whole room do almost all of the answering, and where being wrong is punished in front of everyone within one second of committing. In a workplace, that combination selects hard for seniority and for extroversion, and it silences precisely the people the session was booked to include.

Written answers reward conferring. With a shared sheet and a fixed window of sixty to ninety seconds, the question goes to the table rather than to an individual. Somebody says a partial answer, somebody else corrects it, a third person recalls something adjacent, and the team writes one line. The quiet person's contribution survives, because it goes through the pen rather than through the microphone. Nothing is public until the marking, so being wrong costs nothing socially.

The consequence is blunt. If your objective is entertainment and spectacle, buzzers are better and you should use them. If your objective is anything that could honestly be called team building, written answers are the only defensible choice, because a buzzer format does not build anything. It only reveals who was already fast.

The productive middle is to use both, deliberately, in different rounds. Run the bulk of the quiz on written sheets, then use a buzzer mechanism for exactly one segment where the speed itself is the point: a rapid head-to-head between the top two teams, or a sudden-death tie-break. That gives you the television moment without letting the television format govern the whole event.

If you do use buzzers, three rules make them tolerable. Buzz on behalf of the team, not as an individual, and let the team nominate a different buzzer person each round so the role rotates. Impose a penalty for an incorrect buzz, otherwise the correct strategy is to buzz instantly and think afterwards. And give the question to the other teams on a written sheet if the buzz is wrong, so a failed buzz does not simply destroy the question.

Physical buzzer sets are the one piece of kit that is genuinely hard to improvise. Practical substitutes: a coloured card raised, a table bell, a hand on a shared object, or a phone-based response app if your room has reliable wifi and everyone has a device. Raised cards are slower to adjudicate but they cost nothing, and the ambiguity they introduce is small compared with the ambiguity of five people shouting.

Writing a question that has exactly one defensible answer

Nearly every dispute in a quiz traces back to a question that had two answers and an organiser who only thought of one. There is a test that catches most of them before the day.

The adversarial pass. Take each question and try to justify a different answer. Not a silly one, a defensible one. If you can construct a second answer that a reasonable colleague could argue for using ordinary knowledge, the question is broken. Repair it or bin it. Doing this properly takes about as long as writing the questions did, and it is the step everyone skips.

The recurring faults it finds:

Unbounded superlatives. The largest, the first, the fastest, the most, with no scope. Largest by what measure, first in which country, fastest under which conditions. Bound it: the tallest completed building in Singapore by architectural height as of 2026.

Answers that moved. Populations, records, rankings, company sizes, sporting champions. Anchor to a date and say the date out loud in the question.

Category questions with several members. Name the element that does X, when three elements do X. Either add a discriminating clause or ask for any one of them and accept all.

Ambiguous phrasing of the ask. How many countries border Brazil is clean. How many countries are in South America invites an argument about territories. Prefer counted, defined sets.

Questions where the intended answer depends on your source. If two ordinary reference sources disagree, that question is not fit for a quiz, no matter how good it sounds.

Spelling traps. Names transliterated in more than one way. Decide in advance that phonetic spelling is accepted and say so in the brief, then never revisit it.

Two habits make the whole set safer. Write the answer first and the question afterwards, which forces you to notice how many questions could produce that answer. And add a scope clause to every question that could take one, even when it feels heavy-handed, because a clunky question that cannot be argued with is better than an elegant one that can.

The dispute protocol, and how to run it without losing the room

Something will still be wrong. Plan for it as a procedure rather than as a crisis, and announce the procedure in the opening brief so that it reads as fairness rather than as improvisation.

1. Disputes are written, not shouted. A team that wants to challenge an answer writes the question number and its case on the back of the answer sheet. No verbal challenges are entertained during the round, and the host says so once, at the start, and then simply points at the rule.

2. Disputes are heard between rounds, not during. The host reads the challenges while the next round is being set up. This alone removes most of the disruption, because roughly half of all challenges evaporate once the team has written the argument down.

3. The standard is defensibility, not intention. The question is not whether the team guessed what you meant. It is whether their answer is defensible on the wording as read out. If it is, it scores.

4. If two answers are both defensible, both score. Do not split the point and do not hold a vote. Award the point to every team that gave either answer, including teams that did not challenge but wrote the same thing, which requires the host to re-check the sheets and is the reason this is done between rounds.

5. If the question is simply broken, void it for everyone. Remove the point from all teams, including the team that got it right, and adjust the round maximum. Voiding is cleaner than compensating, and a host who voids one question early in the evening buys enormous credibility for the rest of it.

6. The host decides and the decision is final. Say this in the brief. A quiz with an appeals process has no ending.

One additional protection: nominate a single named adjudicator who is not the host, ideally someone senior enough that nobody argues and detached enough that nobody suspects them. The host runs the room; the adjudicator settles the two or three genuinely contested points. Separating those roles keeps the host in the entertainment business, which is where they are useful.

Scoring, tie-breaks, and stopping a runaway leader from ending the contest early

A quiz has a scoring problem that a physical station event does not. Every team faces the same questions, so scores converge. A typical corporate room finishes five rounds with all teams inside a five or six point band on a sixty point base. That sounds healthy and it hides a real risk: a single team that is genuinely stronger will open a gap in round one and never close it, and by round four everyone else has worked out they cannot catch up.

Three devices, used together.

Cap the contribution of any one round. No round should be worth more than about a fifth of the base, excluding the wager. If your picture round has forty images and your company round has eight questions, the picture round has quietly become the quiz.

Withhold the running total. Read out the round scores, not the cumulative standings, until after round four. Teams need feedback on how they did; they do not need proof that they are out of it. Announcing full standings after every round is the most common self-inflicted wound in this format.

Put a wager round at the end. This is the real answer and it is what makes the last third matter. Teams stake points on their own answers, so a trailing team can gamble its way back and a leading team has to choose between defending and extending.

How to build the wager round

Five questions, and each team receives a stake card with five values on it: 1, 2, 3, 4 and 5. Before each question is read, or better, before the round begins for all five questions at once, teams allocate one stake to each question. A correct answer gains the stake; an incorrect answer loses it. Maximum swing is plus or minus 15.

Allocating all five stakes before hearing any of the questions is the stronger version, because it forces a genuine decision under uncertainty. Teams get the five category headings only, and must guess where their own strength lies. That is a small, clean planning exercise embedded in a quiz, and it is far more interesting than deciding one question at a time with full information.

Sizing the wager. The swing has to be big enough to overturn the leader's margin and small enough that five rounds of work still count. The rule that works: maximum positive swing should sit between the first-to-second gap after the penultimate round and about a quarter of the total points available. With a 60 point base and a typical five point gap, a 15 point maximum swing satisfies both. If you find yourself designing a wager worth 40 points, you have not designed a wager, you have designed a quiz with one round and five hours of preamble.

Tie-breaks. Never leave this to the day. Put the tie-break in the opening brief, in this order: 1. A sudden-death estimation question, closest answer wins, handed to the tied teams on paper simultaneously. This is the best tie-break in existence for a quiz because it cannot be a draw and it cannot be looked up. 2. If somehow still tied, the higher score in the company round, which rewards the thing you actually wanted to reward. 3. If still tied, declare a shared win. Nobody has ever regretted two teams winning.

Two scoring formats worth knowing. The default above is cumulative points with a wager finish. The alternative, for a longer or larger event, is a two-stage format: all teams play four rounds, the top two on points go into a head-to-head buzzer final in front of the room, and the final decides the title outright. It produces a much better spectacle and a much worse participation curve, since most of the room becomes an audience for the last twenty minutes. Choose it only if the event is as much a show as a session.

Six rounds, five teams, ninety minutes: the full point schedule

Thirty people, five teams of six, one host, one marker, one adjudicator, ninety minutes from first word to prize. Written answers throughout, with buzzers reserved for the tie-break if one occurs.

The clock.

Marking happens continuously. Sheets are collected at the end of each round and swapped between teams for marking during the following round's reading, which costs nothing in time and removes the marking bottleneck entirely. The host reads the answers to the previous round while the next round's sheets are being distributed.

Round 1, Openers. Ten questions, 1 point each, maximum 10. Deliberately easy, deliberately broad, deliberately fast. Its job is to get every team writing something in the first two minutes and to confirm that the pens work. Expect most teams to score 7 or higher. A round one that separates teams has failed at its actual purpose, which is settling the room.

Round 2, The House Round. Ten questions on the company, 1 point each, maximum 10.Company history, office trivia, product facts, the number of desks on the fourth floor. This is where you spend your writing effort. It is also the round that makes the quiz feel bespoke rather than bought, and the one people quote back to you weeks later.

Round 3, Faces and Places. Twelve items, 1 point each, maximum 12. Six colleague baby photos with consent obtained in writing, and six close-cropped photographs of locations inside the building. It is the loudest round of the night, and it is placed third because the room needs a noise peak before it can take a hard round.

Round 4, Estimation. Six questions, ranked 3 or 2 or 1, maximum 18. Two questions about the company or the building, four about the world, spanning at least three orders of magnitude. Ninety seconds each, because teams need the time to argue. This is the heaviest round in the base and that is intentional: it is the round with the widest participation.

Round 5, The Hard Round. Ten questions, 1 point each, maximum 10. The general knowledge round, placed fifth and capped at a sixth of the base. Placing it here is deliberate. Early, it demoralises. Last, it decides the event on the one dimension you were trying to demote. Fifth, it creates a visible shake-up in the standings with one round still to play, which is exactly the shape a contest wants.

Round 6, The Wager. Five questions, stakes of 1 to 5 allocated in advance, swing of plus or minus 15. Categories are announced, stake cards are filled in and collected before the first question is read. Maximum base is 60, so a 15 point swing is a quarter of the game.

How it plays out. The following is a representative run, not a real one, and its purpose is to show the shape.

Read the cumulative column instead and the design becomes visible. After round 1, C leads on 9. After round 2, B leads on 16. After round 3, D leads on 25. After round 4, E leads on 34. After round 5, E still leads on 40, with B and C tied on 36, A on 35 and D on 37. Four different teams have led, and going into the final round the whole field is inside five points.

Then the wager does its work. E, leading, allocates its high stakes to the categories it feels safest in and gets two of five right for a net gain of 3, finishing on 43. A, in last place, puts its 5 and its 4 on the two categories it believes it owns, gets three of five right and gains 9, finishing on 44. D does the same and finishes on 46 to win, having led once, in round three, and never since.

The instructive part is the losing margin. C scored the most points of any team across rounds 1, 3 and 5, the three conventional knowledge rounds, and finished last, because it finished last in estimation and gained only 1 net point from a badly allocated wager. That is a quiz working correctly: recall alone did not win it, and the two rounds that required the team to aggregate and to plan did.

Had B and D both finished on 46, the tie-break announced in the opening brief would have applied: one estimation question, both teams writing simultaneously, closest number wins.

Headcount, team composition and how to divide a room of thirty

The floor is about 12 people in three teams. Two teams is a duel rather than a contest, and with two teams the wager round has no field to move through. Local suppliers publish minimums around 20 participants, which reflects what makes commercial sense for them rather than what makes the format work.

The ceiling is where marking breaks, not where the room fills. A single host with written answer sheets can comfortably run 8 to 10 teams. Local specialists publish capacity for 30 to 40 teams and rooms above 200 people, but that requires either a marking team or a digital answer system. If you are running it yourself, cap at 10 teams and add a second marker above six.

Team size is four to six, and five is the target. Below four, one absentee guts the team. Above six, the far end of the table stops contributing, because a written answer sheet has one pen and a table of eight has a physical geometry problem before it has a social one. If you must run teams of eight, use two sheets per team and split the round, or the format quietly reverts to three people playing and five watching.

Composition. Mix departments, mix tenure, and mix age deliberately, because every one of the good question categories rewards a different one of those. A team of five people who all joined in the same year will be strong on the house round and weak everywhere else. Do not let people self-select; pre-assign teams and print the lists.

Seniority. Avoid putting a manager on a team with their own direct reports. In a written-answer format the pen goes to whoever the table defers to, and if that is a person's boss the team stops aggregating and starts agreeing.

The table matters more than people expect. Teams need to be able to lean in and speak quietly. Round tables of five or six work; a long banquet table of ten does not, no matter how you assign the teams sitting at it.

What it costs and how long to book for in Singapore

Published rates. For packaged corporate quiz and game show products, published local figures run from about $20 to $45 per person for online editions and $25 to $50 per person for indoor in-person sessions of 30 to 90 minutes. Half-day formats with facilitation and light refreshments publish in a $45 to $90 per person band, and one local price guide notes an effective rate near $83 per person for a group of 12 because of a $1,000 minimum event fee, falling to roughly $35 per person at 30 people for an in-office session. Several dedicated quizmasters publish no rate at all.

Watch the pricing shape. A quizmaster's real cost is question-writing time plus one evening, and that cost barely moves between 30 people and 100. A per-pax quote for a quiz is therefore a pricing choice rather than a cost structure, and a flat event fee is usually the better deal above about 40 people and the worse deal below 20. Ask for both.

Add-ons published locally. External venue hire adds $10 to $40 per person, function rooms and community centres publish at $200 to $800 for a half day, co-working spaces at $500 to $2,000, hotel ballrooms from $3,000 upward. Weekend surcharges of 15 to 25 per cent and public holiday surcharges around 30 per cent are commonly published, and catering folded into an all-inclusive package typically carries a 30 to 40 per cent markup.

Duration. Ninety minutes is the published local standard and it is correct. Sixty minutes is achievable by cutting to four rounds and dropping the estimation round to four questions, though cutting estimation is cutting the best part. Two hours works if you add a food break between rounds three and four rather than adding rounds, because attention in a quiz decays sharply after about six rounds regardless of how good the questions are. Anything above two hours should be two sessions.

Writing time, if you build it yourself. Budget one and a half to two full days for a six-round set: roughly half a day for the house round, which requires actual research inside your own organisation, half a day for everything else, and a further half day for the adversarial pass and the slides. The adversarial pass is the half day people cut, and it is the half day that prevents the disputes.

Kit list, and which parts cannot be improvised

Must be arranged and cannot be borrowed from a stationery cupboard:

• A PA system and a microphone. Local suppliers publish this as a venue requirement rather than something they bring. A room of 30 needs it more than you think, because a quiz host has to be audible over five tables all talking at once, and a host who has to shout loses their voice by round four.

• A projector or large screen, if you are running picture, ordering or estimation rounds. Reading a picture round aloud is not possible and reading a list of five items for an ordering question aloud is painful.

• A visible scoreboard. A projected spreadsheet is best, because it lets you show round scores without showing the cumulative total until you choose to.

Can be self-provided:

• Answer sheets, printed, one per team per round, with the round name, the team name and numbered lines. Numbered lines matter: they eliminate the entire class of dispute where a team answered question 6 on line 5.

• Stake cards for the wager round. Five cards per team marked 1 to 5, or one printed grid.

• Pens, two per team, and spares. This is the single most common thing to run short of.

• A timer visible to the room.

• Prizes. Small and several beats large and singular, and a wooden spoon for last place is worth more to the room than the winners' prize is.

Optional and genuinely expensive: a buzzer set. Given the argument above, most organisers should skip it and use raised cards for the tie-break.

Room. Round tables, spaced far enough apart that a table conferring at normal volume is not audible to its neighbour. That spacing is the only real constraint on the venue, and it is why a quiz for 30 needs a room sized for 50.

Four ways a company quiz falls apart, and how to prevent each

One round quietly becomes the whole quiz. A picture round of 30 images sits alongside rounds worth 10 points each, and the event is decided by whoever recognises logos. Fix: write the point maximum for every round on a sheet before you write a single question, keep every round inside a fifth of the base, and normalise round lengths rather than question counts.

The same team leads from round one and everyone else disengages. Visible by round three, and the room's energy tells you before the scoreboard does. Fix: withhold cumulative standings until round four, and make sure the wager swing is large enough to overturn the gap. If you can see it happening live, you can also add a bonus to the final round on the spot, but it is better to have designed it correctly.

A question turns out to have two answers and the host improvises. The improvised ruling is always inconsistent with the ruling twenty minutes later, and the room notices. Fix: run the adversarial pass beforehand, announce the written dispute protocol in the brief, appoint an adjudicator who is not the host, and void rather than compensate.

Two people answer everything at each table. The most damaging failure because it is invisible from the front of the room, which looks like five engaged tables. Fix: written answers rather than buzzers, an estimation round where everyone has a number, at least one round with rotating scribes, and a house round where expertise sits with tenure rather than with reading habits. If you want to measure it, ask each table at the end to note who wrote the answers, and read the pattern rather than announcing it.

A fifth, worth naming: the host reads at the wrong speed. Too fast and half the room misses the question; too slow and the energy collapses. Every question should be read twice, in full, with the second reading at the same pace as the first, and the round window announced before the first question rather than after it.

Making it easier or harder by moving one variable at a time

Easier. Extend the answer window to 90 seconds per question. Shorten each round to eight questions. Increase the proportion of recognition and estimation questions and cut the general knowledge round to five questions. Accept partial answers, such as a surname without a first name. Offer one joker per team that doubles any single round, which helps trailing teams more than leading ones.

Standard. As set out in the worked example above.

Harder. Cut the answer window to 45 seconds. Add a connections round in which the link between five answers is itself worth points. Introduce negative marking on the hard round, at minus 1 for a wrong answer and 0 for a blank, which forces teams to decide whether they actually know. Run the wager with stakes allocated blind, before category names are revealed. Add a sudden-death buzzer final between the top two teams.

The single most effective dial in either direction is the answer window, not the questions. Halving the window makes an easy question hard, because it removes the conferring time, and doubling it makes a hard question tractable, because a team that reasons together will get to more answers than any individual in the room would.

What conferring under time pressure actually builds, and what it cannot

What it builds, honestly.

The written-answer version of this format trains one thing well: aggregating partial and conflicting opinions into a single committed answer under mild time pressure. That is a real workplace capability and it is not trivial. A team of six with a ninety second window and five different guesses has to decide how to decide, dozens of times in an evening, with immediate feedback on whether the method worked. The estimation round makes this explicit, because the arithmetic of averaging beats the confidence of individuals often enough for the group to notice.

It also does two smaller things well. It surfaces distributed expertise, in that the house round and the recognition rounds reveal that knowledge in the organisation sits in places the org chart does not mark. And it is genuinely low-barrier: it requires no physical capability, no change of clothes, no fitness, and it works for participants with mobility limitations, for people who dislike physical games, and for a room with a wide age spread, which is a shorter list of exclusions than almost any other team building format.

What it does not build, equally honestly.

It builds nothing physical. Obviously, but worth saying, because a quiz is often booked as the accessible option and then expected to do the job of an active one.

More importantly, it does not require a team to act together, only to decide together. There is no execution phase. No task is divided, no interdependence exists, nobody has to rely on a colleague to carry out a step correctly. A quiz team is a committee producing an output, not a unit performing a task. Formats built around physical stations, timed builds, route navigation or split information force coordination in a way that no question set can, because the difficulty in those formats lies in doing rather than in knowing.

Nor does it build trust in any meaningful sense. Trust develops when someone depends on a colleague and the colleague delivers. Writing down a number nobody was certain of does not create dependency, it creates agreement. Those feel similar in the room and they are not the same thing.

And it does not build communication skills beyond the narrow band of stating a view quickly at a table of six. It exercises no listening under stress, no instruction-giving, no handover, no negotiation with anyone outside your own team.

If the brief you were given is inclusion, mixing across departments, or giving a room something to do together that everybody can participate in, this format delivers. If the brief is collaboration under pressure, trust, or getting a team to execute together, book something else and be honest with whoever asked.

Fifteen minutes of questions worth asking once the scores are read

Do this before people leave the room, while the answer sheets are still on the tables. The sheets are the evidence and they make the conversation concrete rather than sentimental.

On how the team decided:

• Look at the estimation round. How did you settle on the number you wrote? Did anyone suggest averaging, and if not, why not?

• Was there a question where someone at the table was right and the team wrote something else? What happened in that thirty seconds?

• Who held the pen, and how did that get decided? Did it change during the evening?

On who contributed:

• Which round did you personally contribute most to? Was it the round you expected before we started?

• Was there a moment you knew something and did not say it? What stopped you?

• Name one thing you learned about a colleague's knowledge that you did not know at six o'clock.

On the wager:

• You allocated your stakes before hearing the questions. What was the reasoning, and did it survive contact with the actual questions?

• Did the team argue about the allocation, or did one person decide it? Which of those would you do next time?

On transferring it:

• Where in your actual work do you have to combine several people's uncertain estimates into one number that gets committed to? How does your team do that today?

• If we ran this again tomorrow with different questions, what would you change about how your table worked?

Two facilitation notes. Ask these table by table, not to the whole room, because the whole room produces one confident answer from one confident person, which is the exact failure the format is supposed to expose. And avoid asking anything that invites a team to explain why it lost; the scoreboard is not the subject of the debrief and the sheets are.

Aliases, near neighbours, and four questions that expose a thin package

The same format is sold locally under quiz night, pub quiz, trivia night, corporate trivia, game show, virtual game show, trivia challenge and half a dozen supplier-invented names. As with most team building formats in Singapore, the number of names in circulation is roughly double the number of distinct mechanics, and most of the extra names are single-supplier product branding rather than different products. Buy on mechanics, not on names.

What it is not.

Parallel station events such as Mini Olympics or telematch. Teams rotate physically between simultaneous scored activities. No host controls the sequence, the difficulty is in the doing, and the scoring problem is normalising results measured in different units.

Route-based formats such as an Amazing Race. Teams move between locations solving clues in sequence. The design work is in the route and the clue chain, not in a question set, and teams are apart from each other rather than in one room.

Escape rooms. Gated. Solving puzzle one produces the key to puzzle two, progress is mechanical, and time rather than points is the score.

Murder mystery and investigation formats. One long deduction over a fixed cast of suspects, with evidence split between teams. There is no round structure of independent questions and there is only one answer for the whole event.

Physical game show formats. Survey-answer rounds, category boards, stunt segments. These are genuinely different: they need a stage, props and space, and they build closer to what a station event builds. If the quote mentions set-building, this is what you are buying.

Scavenger hunts. Collection and tally, no host-controlled sequence, no closed question set.

The four questions that reveal what you are actually being quoted for.

1. How many of the questions are written for us, and who writes them? The honest answer is usually none, or a company round of ten. That is fine, but it should be priced as fine. A supplier who will build a full house round is doing several hours of research inside your organisation and should charge for it.

2. Buzzers or written answers, and can we choose? If the supplier only owns one mode, you are choosing your event's whole social dynamic based on their inventory. Ask before you book.

3. What is the round structure, and is there a wager or a double-points round? A supplier who cannot describe the round order and the point maximum per round has a slide deck rather than a design. One local supplier publishes a six-round structure with a wagering round, which is the right answer.

4. What happens when a question is disputed? Any answer other than a described procedure means the answer is improvisation.

A fifth, if the quote is per head: what does the price do at 60 people instead of 30? For a format where the marginal cost of an extra participant is one answer sheet, the answer to that question tells you a great deal about how the quote was constructed.

Does PLAYON run game shows?

No. PLAYON is an indoor attractions venue built around physical, self-scoring activities, and a quiz needs seated tables, a controlled sound environment, a host with a microphone and quiet enough conditions that five teams can confer without hearing each other. None of that describes this venue and it would be misleading to pretend otherwise.

Two things are worth saying to an organiser who has read this far.

If the brief that brought you here is genuinely a quiz, build it or buy it from a quizmaster. The design work on this page is the whole job, and no venue substitutes for it.

If the brief was looser than that, and the actual requirement was indoors, air-conditioned, two hours, thirty people, a clear winner and everybody participating, then a scored station rotation answers the same brief by different means. It builds different things, chiefly execution rather than deliberation, and it feels nothing like a quiz. That is a genuine alternative rather than a replacement.

The two also pair well across a longer programme, in this order: physical stations first, quiz afterwards, over food. A quiz works when people are seated and tired, which is exactly the state a station event leaves them in. Running it the other way round wastes both.

What organisers ask before committing to a quiz format

How much does a corporate game show or quiz cost in Singapore? Published local rates run roughly $20 to $50 per person, with online editions from $20 to $45 and indoor in-person sessions from $25 to $50. Several specialist quizmasters publish no rate and quote a flat event fee instead of a per-head price.

How long should a corporate quiz run? Ninety minutes, which is the published local standard. That fits six rounds of eight to twelve questions plus briefing, marking and prizes. Sixty minutes works at four rounds, and anything beyond two hours should be split into two separate sessions.

How many people do you need for a company quiz? Twelve is the working floor, giving three teams of four. Local suppliers publish minimums nearer 20 participants. One host with paper answer sheets handles up to about ten teams comfortably, and above six teams you should add a second marker.

What is the best team size for a corporate quiz? Five people. Four leaves a team exposed to one absentee, and above six the far end of the table stops contributing because a single answer sheet has only one pen and a physical reach limit that shapes who speaks.

Should a corporate quiz use buzzers or written answers? Written answers, for almost every corporate purpose. Buzzers reward speed and confidence, so a handful of individuals answer everything. Written sheets send the question to the table, which is the only version that produces conferring rather than individual recall.

Why do general knowledge quizzes always seem to be won by the same people? Because general knowledge is a lookup task. The fact is either in someone's memory or it is not, so a team can only poll its members and take the confident answer. No conferring occurs, and the best-informed individual effectively plays alone.

What is an estimation round in a quiz? A round where every answer is a number nobody knows exactly, such as a building's height or the office coffee consumption, scored by how close each team gets. Everyone can contribute a guess, so teams must aggregate opinions rather than defer.

How do you stop one team running away with a company quiz? Three measures together. Cap any round at a fifth of the base score, withhold cumulative standings until the fourth round, and finish with a wager round whose maximum swing exceeds the leader's margin but stays near a quarter of the total points.

How does a wager round work in a quiz? Teams receive stake values of 1 to 5 and allocate one to each of the final five questions, ideally before hearing them. A correct answer gains that stake and a wrong answer loses it, producing a maximum swing of 15 points.

How do you write a quiz question with only one correct answer? Write the answer first, then the question, then try to argue for a different answer using ordinary knowledge. Bind every superlative with a measure, anchor anything that changes to a stated date, and discard questions where reference sources disagree.

What do you do when a team disputes an answer? Take disputes in writing on the back of the answer sheet, hear them between rounds rather than during, and award the point to every team giving a defensible answer. If the question is genuinely broken, void it for all teams and adjust the round maximum.

Are guess-the-baby-photo rounds a good idea at work? Yes, with consent obtained individually and in advance from every person whose photo is used. They remove studied knowledge entirely, cost nobody status when answered wrongly, and reliably produce the noisiest and most social round of the evening.

Does a corporate quiz actually build teamwork? Partly. With written answers it genuinely trains aggregating conflicting opinions into one committed decision under time pressure. It does not build trust or coordination, because a quiz team decides together rather than acting together, and no task is ever divided.

Can a company write its own quiz instead of hiring a quizmaster? Yes, and it produces a better event if the company round is properly researched. Budget one and a half to two days of one person's time, including half a day for the adversarial pass that finds questions with two defensible answers.