A session that is one long explanation loses the room in ten minutes. The working shape alternates talking, doing and playing - roughly in that proportion.
Two rules that matter more than the timings: never explain for more than fifteen minutes, and always finish with play. A child should leave having played, not having listened.
A year is roughly thirty sessions. A workable distribution for a beginners' group:
Tournaments are woven through the year rather than saved for the end - see below. And the programme is a guide, not an obligation: if the group has not absorbed mates by session twelve, do not move on because the plan says so.
A group that never competes stops being interesting around the third month. Events are what give the learning a point.
Running an event with children - a school tournament, what to announce before round one.
The standard problem: half the group has been coming for two years, half arrived in September. Splitting is not always possible, so the session has to absorb the difference.
The strongest predictor of a child staying is not talent but whether they have someone to talk to in the room. Which is why the first session for a newcomer should end with them having played somebody and learned a name - what chess actually gives children.
Less than people assume in terms of playing strength, and more in terms of everything else.
What to buy and in what order - club economics; how to start the whole thing - starting a club.
Alternate talking, solving and playing: a short warm-up puzzle, ten to fifteen minutes of new material, ten to fifteen minutes of puzzles on the same theme, then twenty to thirty minutes of actual games with clocks, and a brief wrap-up. Never explain for more than fifteen minutes, and always finish with play.
Roughly thirty sessions: rules with no gaps first, then basic checkmates, then tactics one device at a time, then three opening principles, then simple endgames. Internal tournaments run through the year rather than being saved for the end.
Aim explanations at the middle, prepare puzzles in two difficulties on the same theme, let stronger children explain things to newcomers, use handicaps in games and pair beginners with beginners for most of the playing time.
Make sure they play every session, that beginners are paired so nobody loses ten games in a row, that progress is visible through a rating or a wall table, and that mistakes are never mocked. Children stay for the company as much as for the chess.
For beginners, a confident club player is enough - the ability to explain matters far more than playing strength. What you do need is a demonstration board, enough sets for everyone to play at once, clocks for half the tables and prepared puzzles.
Pairings are computed by bbpPairings, an open-source engine implementing the Dutch system by FIDE rules: score groups, floats on an odd field, colour alternation. Free to use.
The bot is constantly improved based on real tournaments. If you need a feature that isn't there yet, message the developers right inside the bot with the /support command. Many features appeared exactly because organizers asked for them, and useful ideas get added quickly.
For organizers - starting a club, school tournaments, club economics. For parents - what age to start.