Key takeaways
- →Self-determination theory says motivation depends on three needs — autonomy, competence and relatedness — and a session can starve any of them by accident.
- →Autonomy is usually the cheapest to restore: give people a genuine choice about order, question or partner rather than a choice about nothing.
- →Competence collapses when the task is either trivially easy or impossible to attempt, so pitch activities where about two thirds of the room can make visible progress.
- →Relatedness is the need remote sessions starve first, and it is rebuilt by small stable groups rather than by whole-room sharing.
- →Points and leaderboards support motivation when they mark progress and undermine it when they replace the reason for doing the work.
- →Diagnose before you redesign: low response rates, hedged answers and silent breakouts each point at a different missing need.
Self-determination theory is the most useful piece of motivation research a facilitator can carry into a room, because it names the three things people need in order to engage with anything: autonomy, competence and relatedness. When a session goes flat, it is almost always because your design quietly removed one of them.
The theory, developed by Edward Deci and Richard Ryan, distinguishes between doing something because it matters to you and doing it because someone is watching. Both produce compliance. Only the first produces the kind of contribution you actually wanted when you booked ninety minutes of eleven people's time.
This article is about the operational version. What each need looks like in a workshop, which one your room is missing, and the specific design changes that restore it — timings, group sizes, question wording, and the two places where points and leaderboards help rather than hurt.
What the theory actually claims
Self-determination theory separates motivation by quality rather than quantity. Someone can be highly motivated to finish your exercise purely because their director is in the room, and that motivation will evaporate the moment the director leaves. Someone else can be quietly, durably interested in the problem. The behaviour looks similar for the first ten minutes and diverges completely after that.
The three needs are not preferences. The claim is that they function like nutrients: satisfy them and self-motivation follows more or less on its own; frustrate them and you get withdrawal, minimal compliance, or the resentful version of participation where people answer in the fewest possible words.
For facilitation this reframes the problem. You are not trying to motivate a room, which you cannot do from the front. You are trying to remove the things in your own agenda that are getting in the way.
Autonomy: give a choice that is actually a choice
Autonomy is the need most often faked. Asking a room 'shall we start with the roadmap or the hiring plan?' when you have already built forty slides on the roadmap is not a choice, and people recognise it instantly. A fake choice costs you more credibility than no choice at all.
Real autonomy in a session is narrow and structural. Which of three prompts you answer. Whether you contribute by voice or in writing. Whether your response carries your name. Which two people you spend the next eight minutes with. None of these threaten your agenda, and each one converts a passive attendee into someone who has made a decision about their own participation.
The wording matters as much as the option. 'You have four minutes — spend them however is useful' supports autonomy. 'You have four minutes, and I want everyone contributing at least two ideas' removes it, and the ideas you get back will be padded to hit two.
A specific swap that works in nearly any session: replace one round of whole-group questions with an open-response activity where people choose which of three prompts to answer. In Session Flo this runs as a single open-text or word cloud activity with the prompts on screen, and the spread of which prompt people picked tells you where the room's attention actually is.
Competence: pitch the task where effort shows
Competence is not confidence and it is not expertise. It is the felt sense that what you are doing is producing progress. That sense dies in two directions: tasks so easy that finishing them proves nothing, and tasks so ambiguous that you cannot tell whether you have started.
The practical target is an activity where roughly two thirds of the room can make visible progress in the time given. Below that and people stall and go quiet; well above it and the exercise reads as busywork. For a quiz, that means a question set where the median score lands somewhere near sixty to seventy per cent rather than a set where everyone scores ten out of ten.
Feedback speed does most of the work here. A quiz question answered and scored within fifteen seconds gives a clean signal; the same question marked at the end of the session gives almost none. This is the same mechanism behind retrieval practice — the attempt plus the immediate correction is what builds both the memory and the sense of getting somewhere.
The most common competence failure in workshops is the unbounded prompt. 'How should we improve onboarding?' has no visible edge, so nobody knows what a good four minutes looks like. 'Name the one step in onboarding where new starters get stuck most often' has an edge, and people will hit it.
| Need | Frustrating version | Supportive version |
|---|---|---|
| Autonomy | 'Everyone give at least two ideas' | 'Answer whichever of these three prompts is closest to your work' |
| Competence | 'How should we improve onboarding?' | 'Name the one onboarding step where new starters get stuck' |
| Relatedness | 'Turn to whoever is next to you' | 'You are in the same trio all morning — introductions first, thirty seconds each' |
| All three | 'We will go round the room in order' | 'Two minutes silent, then your trio picks one to share' |
Relatedness: the need remote sessions starve first
In a room, relatedness accumulates for free. People arrive early, they talk about the weekend, they sit next to someone they know. On a video call none of that happens: the session begins at the second it begins, with fourteen unrelated rectangles and no side conversation.
The fix is not a longer icebreaker. It is smaller, more stable groupings. A trio that stays together for three separate activities builds more connection than six rotating breakouts, because the second time you are put with the same people you already know something about them and can skip the reintroduction.
Keep breakouts at three to four people for discussion where everyone should speak. At five the quietest person can pass without it being noticed, and at seven or more you are watching the Ringelmann effect operate in real time: per-person effort falls as the group grows, not because anyone decided to coast but because the marginal contribution stops feeling necessary.
One structural detail: name the groups and put the names on screen. 'Group 3' is anonymous. 'The team with Priya, Marcus and Wen' is not, and being individually identifiable within a small group is most of what relatedness needs to get started.
A sixty-minute session designed around the three needs
0–5 min — a low-stakes entry with a real choice
Open with a one-question pulse check where people pick which of two or three framings matches their week. It costs nothing, it gives everyone an early autonomous act, and the distribution on screen tells you what mood you are actually working with before you commit to your first section.
5–12 min — form the trios and let them talk
Assign fixed trios for the whole session and give them ninety seconds of introductions with a specific prompt — not 'introduce yourselves' but 'what are you closest to in this topic?'. Say clearly that these groups stay together, so the investment is worth making.
12–28 min — the competence block
Put the hardest thinking here, while attention is highest. Use a bounded prompt with a visible timer and a defined output: one sentence, one number, one named blocker. Silent individual work first, then trio discussion — the order matters, because talking first anchors the trio on whoever speaks fastest.
28–40 min — surface and compare
Collect the trio outputs into one shared view and read the pattern aloud. This is where competence gets confirmed: people see their contribution land in something bigger. Keep the reveal to what the room submitted rather than adding your own answer on top.
40–52 min — converge with a vote
Run a dot vote or ranking across the collected items. Voting is an autonomy act with real consequence, which is why it lifts engagement in a way that discussion alone does not. Show the result immediately and say out loud what it means for what happens next.
52–60 min — close on a commitment, not a summary
Ask each trio for one thing they will do or check before the next session, in a sentence. A commitment made inside a small group where you are known is materially more likely to survive the week than one made to a facilitator you will not see again.
Where points and leaderboards actually help
Extrinsic motivators are not banned by self-determination theory, but they are conditional. A reward that informs — this is how you did, this is where you improved — supports competence. A reward that controls — do this and you get a prize, fail and everyone sees — pushes motivation outward and tends to shrink effort to the minimum that earns the points.
In practice, gamification works best where the underlying activity is already worth doing and the scoring simply makes progress visible. A quiz in a training session is a good candidate: the retrieval is doing the learning, and the score is feedback. A leaderboard in a psychological-safety discussion is a bad candidate, because the thing you want is candour and the thing you are rewarding is speed.
Two guardrails keep scoring on the useful side. First, score teams rather than individuals when the topic is at all sensitive — it keeps the competitive energy and removes the personal exposure. Second, cap the leaderboard display at the top few and never show a full ranked list of everyone, because the bottom third learns only that participating was a mistake.
✓ Pros
- •Scores that mark progress across rounds, so improvement is visible
- •Team-level points where the topic carries any personal exposure
- •Immediate per-question feedback rather than a total at the end
- •Optional participation in the scored portion, with the content available either way
- •Recognition tied to contribution quality that the room itself voted on
✗ Cons
- •Full ranked lists that name everyone from first to last
- •Prizes announced before the activity, which reframes the work as a means to the prize
- •Points for volume — most ideas, most comments — which reliably produces filler
- •Scoring attached to opinions or feelings rather than to knowledge
- •Leaderboards in sessions whose purpose is surfacing problems
Diagnosing which need is missing
Top Tips
- Response rate under two thirds on a low-effort question usually means autonomy: people have decided this is happening to them rather than with them. Offer a genuine choice in the next activity and watch whether it recovers.
- Answers that are technically correct but three words long point at competence. The task is either too vague to attempt properly or too obvious to bother with, and shortening the prompt will not fix either.
- Breakouts that stay silent for the first ninety seconds are a relatedness problem, not a shyness problem. Groups that have never been introduced spend their first minute deciding who goes first.
- A room that answers you but never each other has all three needs partly met and no structure for peer interaction. Add a paired step before the whole-group step and the pattern changes within one activity.
- Enthusiastic participation that collapses the moment a senior person leaves the call is the clearest signal of controlled rather than autonomous motivation. Nothing in your next slide will fix that; the framing of the whole session needs to change.
- If people ask 'is this being recorded?' or 'will this be shared?' before answering, autonomy over identity is the live concern. Switch the activity to anonymous responses and say so explicitly before the question appears.
The mistakes that cost the most
The first is treating enthusiasm as the goal. High-energy sessions with no autonomy produce a good hour and nothing afterwards, because nobody chose anything they now feel responsible for. Energy is a by-product of the three needs being met, not a substitute for them.
The second is over-choosing. Autonomy support does not mean offering seven options for everything; too many decisions raise cognitive load and people default to whatever is listed first. Two or three meaningful options per activity is the working range.
The third is assuming relatedness carries over between sessions. It decays. A group that worked well together in January will need thirty seconds of reconnection in March, and skipping it because 'they already know each other' is how a recurring workshop slowly turns silent.
The fourth is applying all of this to a session that should not exist. If the meeting has no decision to make and no problem to work on, no amount of need-supportive design will rescue it. Send the update in writing and give the hour back.
You cannot motivate a room from the front. You can only stop removing the three things that let people motivate themselves.
Making it stick across a series
For a recurring session — a fortnightly team workshop, a training programme, a quarterly planning rhythm — the three needs compound. Fixed small groups across sessions build relatedness that no single icebreaker matches. A visible skill or score trajectory across weeks builds competence in a way a one-off quiz cannot. And autonomy grows when participants start choosing agenda items rather than receiving them.
The strongest version of that last point is letting the room set part of the next agenda before this one ends. Collect proposed topics in the final five minutes, vote on them, and open the next session with the winner. It takes six minutes, it removes the agenda-writing problem entirely, and it makes the following session something people opted into.
A short written recap sent afterwards does quiet work on all three. It confirms what the group produced, it lets people who chose not to speak see that their vote counted, and it makes the next invitation easier to accept.
Frequently asked questions
What are the three needs in self-determination theory?
Autonomy, competence and relatedness. Autonomy is having genuine say over how you take part; competence is the felt sense that effort produces visible progress; relatedness is being known by and mattering to at least a few people present. The theory holds that motivation is highest and most durable when all three are met, and that frustrating any one of them produces withdrawal or minimal compliance.
Does this mean I should never use points or leaderboards?
No. Extrinsic rewards support motivation when they are informational — when the score tells you how you are doing and where you improved — and undermine it when they are controlling, meaning the reward becomes the reason for the activity. A quiz score in a training session usually informs. A prize announced beforehand for the most ideas usually controls, and produces padded ideas.
How do I support autonomy without losing control of the agenda?
Offer choices that are narrow and structural rather than choices about the agenda itself. Which of three prompts someone answers, whether they respond by voice or in writing, whether their name is attached, which small group they work in. Each is a real decision about participation, none of them threaten your timings, and two or three per session is enough.
What is the fastest way to build relatedness on a video call?
Fixed trios with a specific introduction prompt, reused across the whole session. Ninety seconds of 'what are you closest to in this topic?' in a group of three that stays together does more than a fifteen-minute whole-group icebreaker, because the connection gets used again rather than being made once and abandoned.
How do I tell whether the problem is motivation or something else?
Look at what happens under low effort. If people will not answer a one-tap poll, the barrier is not workload — it is autonomy or trust. If they answer polls but give three-word text responses, it is usually competence: the prompt is too vague to attempt. If they answer you but never each other, you have no peer structure. Each of those has a different fix.
Does self-determination theory apply to mandatory sessions?
Yes, and it matters more there. Attendance being compulsory removes one layer of autonomy, so the autonomy you can offer inside the session carries more weight. Being explicit about it helps: acknowledge that people did not choose to be there, then give them real choices about how they spend the time they are giving you.
Keep reading
- where gamification at work helps and where it hurts — The longer version of the informational-versus-controlling reward distinction.
- why big meetings kill contribution — What happens to relatedness and effort as group size grows.
- getting quiet people to contribute — Autonomy over identity in practice: anonymous responses and written channels.
- using leaderboards without damaging team morale — Concrete rules for capping and framing a scoreboard.
- Session Flo plans — For teams running recurring sessions with fixed groups and saved activity sets.