Clever Groups
  • Try It
  • How It Works
  • Blog
  • What's New
  • Pricing
  • Guest
  • Register
  • Login

How Clever Groups Works

Clever Groups is used by teachers, corporate team leaders, camp coordinators, and many others. This guide describes things from a teacher and their class of students perspective — the same concepts apply regardless of your context, and can handle up to 1,000 team members.

There are two things to understand: the model (what the pieces are) and the workflow (what you and your students actually do).

1 hour 10 hrs 100 hrs 1,000 hrs 50 100 150 200 time by hand people by hand

A good guess vs. a confirmed best

You can do this by hand — teachers do it every year for the annual school camp. Open a spreadsheet, spend a weekend on it, open a bottle of red. You finish on Sunday night feeling good about it. Then Monday arrives: the students who missed out say so, the parents say so louder, and you realise you were working from instinct, not from a confirmed best arrangement.

Preference Groups replaces the weekend with seconds. Demand Groups goes further: rather than maximizing the overall score, it works to ensure every individual ends up with at least one person they chose — a personal win for each person, not just the best average. The number of possible arrangements grows so fast with class size that beyond about 60–80 students, no human can work through them all. That’s where Clever Groups earns its place.

100 1K 10K 100K 1M 10M 100M Random · ~500 PG · ~10M DG · ~22M

preference checks per generation · class of 30 students · log scale

Random Groups
Equivalent to
~500
preference checks
per generation · class of 30
Preference Groups
~10 million
preference checks
per generation · class of 30
Demand Groups
~22 million
preference checks
per generation · class of 30

Each check asks: is this student with someone they prefer — and would swapping them produce a better arrangement? Demand Groups runs harder than Preference Groups because it works to ensure every student ends up with at least one person they chose. Random Groups places students at random and then scores the result, but skips the optimisation loops — a few hundred checks vs millions.

Part 1 — The Model

Three things go in. One thing comes out.

1

Your people — Team Members

Teacher adds these

A Team is everyone you want to group together — a single class, multiple classes, a whole year level, a school cohort, or any other group of people. Everyone in it is a Team Member. You can label each person with a category — any label that makes sense for your context: Boy / Girl, High / Mid / Low ability, Beginner / Intermediate / Advanced. Clever Groups uses categories to balance group composition.

Categories are optional. Add them when group balance matters; skip them when any mix will do.

You can have up to 20 teams — one per class, year level, subject, cohort, or any other grouping you need.

Boy | Girl
Liam
Aaliyah
Diego
Sofia
Tyler
Maya
Connor
Priya
Jackson
Zara
Marcus
Olivia
2

Preferences — who each person wants to work with

Teacher or students add these

Each Team Member names the classmates they’d like to be grouped with, in order. Preference 1 is the most important — Clever Groups works through them in sequence. There’s no fixed limit on the number of preferences.

In practice, 2–3 preferences is typical. Clever Groups tries to honor as many requests as possible across the whole class — not just the most popular students.

Preferences can be entered by the teacher directly for any student who hasn’t yet joined the team — useful when you want to run the full process yourself, or when students are too young to submit their own. Once a student joins and submits their own preferences, they manage their own data.

Joining and submitting preferences is always free for students — no account, no payment, no friction.

Aaliyah
prefers…
1st
Maya
2nd
Zara
• • •
3

Constraints — rules the groups must follow

Teacher adds these

Constraints are optional rules you set before generating. Clever Groups respects them all while still maximizing preference satisfaction.

Keep Apart

Pairs of students who must never be in the same group.

Tyler
×
Connor
Sofia
×
Aaliyah

Group Sizes

Specify the number of groups of each size — e.g. two groups of 3 and one group of 4. Groups can be different sizes.

Group of 3
Group of 4
Group of 6

Group Composition

The balance of categories within each group — e.g. 2 boys + 2 girls per group.

Group of 4
2 boys + 2 girls
Group of 2
1 boy + 1 girl
Part 2 — The Workflow

What you do, and optionally what your students do.

Two ways to collect preferences: Share the hash code and let students submit their own — or enter preferences yourself on the teacher dashboard for any student who hasn’t joined yet. If you’re entering preferences yourself, skip the Students column below and Steps 2 and 5.
You (Teacher)
1
Create a team Give it a name — your class, club, or group. A unique hash code is generated automatically. Optionally, enter or upload your student list upfront and assign categories (e.g. Boy, Girl, Learning Support, Class Leader) before students submit preferences — or let students enter their own names when they join. Your choice.
2
Share the hash code Copy the share link to your clipboard and paste it into an email — students who click it arrive at the home page with the code already filled in. Or simply read the code out to your class. Skip this step if you’re entering all preferences yourself — the hash code is only needed for student self-submission.
—
—
5
Chase up anyone on blue Each student has a status button on your dashboard. A document icon means they haven’t joined yet. Once they join it becomes an avatar — blue while preferences are outstanding, green once submitted. Chase up anyone still showing blue. Only applicable when students are submitting their own preferences — skip if you’re entering preferences yourself.
6
Set constraints In the Student Distribution dialog on the teacher dashboard: add any keep-apart pairs first, then set your Group Entries (e.g. 2× groups of 3 + 1× group of 4), then optionally add Category Entries for gender or ability balance.
7
Generate groups Choose your generation mode — Preference Groups (default, seconds), Demand Groups (works to place every person with at least one person they chose — or confirms it’s impossible, up to 3 minutes), or Random Groups — then hit Create. See below for the difference.
8
Review, tweak & save Check the satisfaction score. Drag and drop students between groups to tweak manually. Display, export or print your groups. Save to your group history and compare satisfaction scores across generations.
Your Students (optional)
—
—
3
Join the team Click the link from the teacher’s email — the hash code is pre-filled, just enter a name and hit Create/Join. Or go to clevergroups.com and enter the code manually. No student account needed.
4
Submit preferences Choose who you’d like to work with, in order. First preference counts most. Takes about a minute.
—
—
—
—
Three ways to generate groups

Everyone submits the same prioritized list of who they’d like to work with. The generation mode determines what Clever Groups does with those lists.

Preference Groups

The default mode. Clever Groups works through each person’s list in priority order, placing people into groups that honor as many preferences as possible across the whole team.

Someone may not get their top preference if placing them differently creates a better outcome for everyone else. Clever Groups is always weighing the individual against the whole — finding the arrangement that does the most good across the class.

Students who didn’t submit preferences are placed at random, so each run produces a naturally varied arrangement — no separate randomise option needed.

✓ Keep-apart pairs respected    ✓ Runs in seconds    ✓ 1 credit per run (free trial & Foundation Free)

Demand Groups

Where Preference Groups maximizes the overall satisfaction score across the groups, Demand Groups treats each preference as a demand — working to ensure every person ends up with at least one person they chose. After generation it tells you exactly what percentage had their first preference placed, and what percentage got any preference placed. If someone can’t be placed with anyone they chose given your group sizes and keep-apart rules, it tells you exactly who and how close it got.

Two sub-options control which preferences count:

All preferences — Clever Groups looks across each person’s full list; more options to work with, and more likely to place everyone with someone they chose. Best when students have submitted 2 or more preferences.

First preference only — focuses exclusively on each person’s top choice; best when only one preference was collected, or when only the top pick matters.

✓ Keep-apart pairs respected    ✓ Runs in seconds    ✓ 1 credit per run (free trial & Foundation Free)

Random Groups

Pure random placement. Every person is assigned to a group without reference to any preferences submitted. Each run produces a very different arrangement.

The satisfaction score and preference counts are still measured afterwards — so you can compare a random run directly against a preference-based one on the same team data.

✓ Keep-apart pairs respected    ✓ Runs in seconds    ✓ 1 credit per run (free trial & Foundation Free)

Part 3 — The Result

Students submit preferences. Clever Groups finds the best arrangement and gives you groups to review, tweak, and apply. Here are four situations teachers commonly face.

Preference Groups
Running a class project
Year 10 science · 30 students · 6 groups of 5 · six-week project
Aa
Aaliyah
Chose Maya as her first preference and Zara as her second. Clever Groups placed her with Maya — her top pick.
First choice ✓
Mc
Marcus
Chose Connor first, Tyler second. Connor’s group was already full when the preference grouping reached Marcus, so he was placed with Tyler instead.
2nd choice ✓
Pr
Priya
Nominated only Aaliyah. One preference was enough — placed together.
First choice ✓
… and 5 more groups
84% overall group satisfaction — the highest arrangement found for this class.
Score reflects rank — first choice counts more than second, getting neither scores zero. Clever Groups finds the arrangement that maximizes the class total. Drag and drop to tweak, then save.
Demand Groups
A camp or excursion — no one should be isolated
Three Year 8 classes · 120 students · overnight camp · cabin groups of 4
Co
Connor
Submitted just one preference — Jackson. One nomination was enough. They ended up in the same cabin.
Same cabin ✓
So
Sofia
Chose Aaliyah first, Maya second. Aaliyah’s cabin was already full, so Sofia was placed with Maya — Sofia still got someone she chose.
Li
Liam
Listed three preferences. Placed with his first pick, Connor.
First choice ✓
… and 29 more cabins
120 of 120 placed with at least one person they chose, in priority order.
At 120 students, no teacher works through every possible arrangement manually. Clever Groups works hard to place every student with at least one person they chose. If it’s ever impossible given your group sizes and keep-apart rules, it tells you exactly which students and how close it got.
Smaller groups (under 15 or so): Preference Groups often produces equal or better results — at small scale the individual-placement objective rarely adds anything over what Preference Groups already achieves. As class size grows and the preference graph becomes more complex, the Demand Groups approach earns its place.
Random Groups
A field day or house event — random and fair
Year 9 · 120 students · annual field day · 4 color teams of 30 · 8 keep-apart pairs
Di
Diego
✕
Ty
Tyler
Flagged as a keep-apart pair by the teacher. Clever Groups ensured they ended up on different teams regardless of the random draw.
Separated ✓
R1
Twin 1
✕
R2
Twin 2
The Ramirez twins — their teacher requested they be on separate teams. Both keep-apart pairs honored without any conflict.
Separated ✓
+
116 more
Everyone else drawn at random — 30 students per team, balanced and done in under a second.
Random draw
… 120 students across 4 teams
120 students, 4 teams. All 8 keep-apart pairs honored. Done in under a second.
When fairness means random, not optimized. Keep-apart pairs are always honored regardless of mode. Re-run any time for a completely new arrangement.
Preference Groups
A new year, a new mix — not everyone knows someone yet
Three Year 7 classes · 60 students · first-week activity · groups of 4

38 of 60 submitted preferences — many are new and don’t know anyone yet.

Aa
Aaliyah
New to the school — knew Emma from her previous school and nominated her. Clever Groups honored it. Placed in the same group.
Together ✓
Za
Zara
Listed Maya and Priya as her preferences. Ended up in the same group as Maya — her first choice honored.
Together ✓
Mc
Marcus
Doesn’t know anyone yet and submitted no preferences. Placed randomly — completely normal at the start of the year.
Placed randomly
… and 11 more groups
38 of 60 submitted preferences. 33 placed with at least one they chose. 22 placed randomly — expected at start of year.
Honors the connections that exist without requiring every student to have them. Preference Groups places students who submitted preferences with someone they chose, while students who didn’t are placed at random — a natural part of how the mode works. Re-run any time for a different arrangement.

Ready to try it?

Start with a free demo — 12 students pre-loaded, no account needed.

Try it free Register — it’s free
Pricing
Terms
Support
Privacy
Contact
Cookies

Clever Groups®. Copyright 2016-2026.
All rights reserved.