How to Run Online Chess Classes That Actually Work

An online chess class is three layers stacked, and a weakness in any one of them quietly undermines the whole thing. The board layer is where chess actually happens: students need to see positions and, critically, move pieces themselves. The video layer is how you teach and read the room: faces, tone, who is lost and who is bored. The admin layer is everything around the class: who attended, who played between sessions, what the parents were told. Coaches obsess over the first layer, tolerate the second, and neglect the third, which is exactly backwards for retention, because the third layer is where students are kept or lost.

The good news is that the tooling decisions for all three layers are quick to make and mostly free to start. You do not need a custom platform or a big budget to run online classes that retain. You need a shared board your students can move on, a reliable video call, and a single place to record the operational facts. The hard part is never the tools; it is the discipline of an interactive format and consistent between-class follow-up, which no tool supplies for you. This guide gives you the checklist for the tools and a template for the discipline.

Use this checklist to assemble your stack. The principle behind every row is the same: the student should be active, the coach should be able to see and verify, and the operational facts should land somewhere durable rather than in a chat that scrolls away. You can start entirely on free tools and add paid ones only where a specific friction justifies it.

A note on the board layer, because it is the one coaches get wrong most often. Screen-sharing your own analysis board feels efficient and is the default everyone reaches for, but it makes the class a broadcast: students watch you move pieces and their hands stay still. A shared interactive board, such as a Lichess study where multiple people can move at once [1], changes the whole dynamic, because now every student can be asked to find the move on their own board and you can see who did. If you choose between accounts and platforms for your students, our comparison of Lichess and Chess.com for academies weighs the trade-offs for group teaching, since Chess.com also runs its own shared analysis tools and support documentation for coaches setting up classes [4]. Whichever platform you land on, both expose a student's recent games through a documented API [2], which is what lets a coach verify between-class play at the account level instead of asking each family to report it.

The format below is the operational heart of this guide. It is built on one rule: never let the class become a lecture, because online a lecture is a stream of half-attention waiting to wander. Every block ends with the students doing something, not just the coach demonstrating. Adapt the minutes to your batch length; keep the rhythm of short, active segments. A puzzle trainer such as Lichess's [3] is a fast way to pull a fresh warm-up set each week without preparing one by hand, which matters because the warm-up only works if it is ready before the first student joins.

1. 1Minutes 0 to 5, warm-up: a puzzle every student solves on their own board while you arrive and take attendance. Class starts active, not waiting.
2. 2Minutes 5 to 8, recap: open someone's between-class game or result from a shared study, so last week's homework visibly mattered.
3. 3Minutes 8 to 20, teach-and-do block one: introduce a concept for a few minutes, then set a position every student solves on their board with cameras on.
4. 4Minutes 20 to 32, teach-and-do block two or paired work: a second concept, or pair students on a shared board so the segment is social and low-stakes.
5. 5Minutes 32 to 42, applied play: a guided game or themed exercise where students apply the day's idea while you watch and prompt.
6. 6Minutes 42 to 45, assign and close: set a few puzzles and at least one game to play, and name one thing each student did well today.

Two structural choices in that template do most of the retention work. The recap at minutes 5 to 8 closes the between-class loop: a child who knows their game might be opened next week actually plays one, which is the behaviour that turns a flat rating into real improvement. The personalised close at the end means every student leaves the call having been seen by name, which is a surprisingly large part of why a child wants to come back. Neither costs you anything but intention, and both are routinely skipped under time pressure.

The class is one thing; running an academy of classes is another, and the operations are where online academies either stay lean or drown in scattered admin. A handful of rules, set early, keep the weekly load to minutes and keep students from leaking through the gaps between tools.

• Stable batches: keep the same students together each week so peer familiarity builds, which it never does if membership churns session to session.
• Cameras-on as a stated expectation set at enrolment, so it is normal rather than a fight, and the coach can actually see engagement.
• One record per student for attendance, synced games, and rating, so a weekly drift check is a glance rather than an investigation across three apps.
• A defined make-up rule for the inevitable missed online class, decided in advance rather than negotiated each time it happens.
• A regular parent update on a fixed cadence, because for an online family the update is the only window they have into the class.

Those rules cover the routine week. The situations that actually test an online academy are the live-session failures that never happen in a physical room, and each is far easier to hold if you decide the response before it happens rather than improvising while a dozen children wait on the call. Five recur often enough to be worth a written rule, and none of them should cost the coach a fresh decision each time it occurs.

• A student's connection drops mid-session: do not mark them absent if they were present for the opening segment. Drop the saved study link and the block's target position into the chat so they rejoin at the right place or finish it offline, and spend no make-up credit on a technical fault the family did not cause.
• A student keeps the camera off: because cameras-on was set as an expectation at enrolment, the first move is a quiet private nudge, not a call-out in front of the group. If the camera stays dark for two sessions running, it becomes a parent conversation, since a face going dark online is the earliest visible sign of a student starting to drift.
• One confident student answers every position out loud: switch that segment to everyone solving on their own board with hands up when they have the move, so the shared board stops being a race won by the fastest voice, then call on a quieter student by name for the next position.
• A young child loses the position on the shared board: keep a coach-owned reset chapter with the clean starting position one click away, and treat resetting as a normal, blameless action rather than a scramble that stalls the whole group while one child catches up.
• A student did not play the assigned game between classes: the synced games reveal it before the recap starts, so adjust on the spot with a short guided game in the applied-play block instead of reviewing a game that does not exist, and send a note home that names the missing game plainly rather than papering over it.

The single biggest operational lever is consolidation. An online academy run across a video tool, a spreadsheet of attendance, the game platform, and several WhatsApp threads forces the coach to assemble a picture of each student by hand every week, which is exactly the work that does not get done when a term gets busy, which is exactly when students start drifting. Keeping attendance, synced games, and parent communication on one student record is what makes the weekly retention scan sustainable. The broader weekly operating rhythm, scheduling, fees, and follow-up, sits in our pillar on running a chess academy and in our guide to academy management software.

Between-class follow-up is where software earns its place in an online academy, because it is the most valuable work and the easiest to drop. The loop is: students play and solve between classes, you verify they did via synced games, and you send parents a short update built from that real activity. A ChessCore rule we hold to is that when the platform drafts those updates, it reports only genuine data from the student's actual games and a coach approves each one before it sends, so the proof an online parent receives is always true. Automation should make the follow-up loop faster to run, never invent results to fill it. Most academies send that update over WhatsApp, the near-universal messaging channel families already open daily [5], rather than a separate portal parents have to remember to log into.