Chess Progress Report: Academy Guide

A chess progress report needs exactly three things: evidence, narrative, and a next step. Evidence is the verifiable record of the month: classes attended, games played, rating movement, and one or two engine-checked moments from the student's own games. Narrative is the coach's reading of that evidence in three or four sentences. The next step is the single thing the family should focus on before the next report. Everything else is decoration that costs writing time and reading attention.

Notice what is missing. No generic paragraph about the benefits of chess. No ten-point skills rubric scored three out of five across the board, which parents learn to ignore by the second month because every child scores three out of five on everything. No promises about future ratings. Parents are busy; the report that survives contact with a working parent's evening is one screen long and answerable in one reply. If you want a ready-made structure to copy, our progress report template walks through a fill-in-the-blanks version of this exact anatomy.

The single next step deserves emphasis because it is the part most academies skip. A report that only describes the past reads like a school report card: noted, filed, forgotten. A report that ends with one action, like fifteen minutes of rook endgame practice twice a week or playing two slow games instead of twenty blitz games, gives the family something to do and gives next month's report its opening line. Did the next step happen? That continuity is what turns twelve isolated documents into a story parents can actually follow.

Monthly reports die because they are the highest-effort, lowest-urgency task in the academy. Nothing breaks on the day a report does not go out. No class is cancelled, no fee goes uncollected, no parent calls to complain, at least not that month. So when the founder is also the head coach, the scheduler, and the fee collector, report writing is the task that always loses the triage. The decay pattern is so consistent you can set a calendar by it: month one ships because enthusiasm is high, month two ships late and shorter, month three quietly does not ship at all.

The root cause is that each report is built from a blank page and from memory. To write one honest report you need to remember or look up how many classes the student attended, find their recent games, recall what you worked on with them, and then compose prose that does not repeat last month's. Multiply that by a batch of fifteen students and you have an evening of unpaid administrative writing for a coach who already taught all week. The marginal cost of every report is as high as the first one, and the academy pays that cost again every single month. Work shaped like that does not survive contact with a busy term, a tournament weekend, or a coach falling ill.

There is a quieter failure mode too: the reports that do ship become vague. When the writing burden is high and the data is scattered, coaches retreat to safe generalities. Aarav is making good progress and should practise tactics. Sentences like that are unfalsifiable, identical across students, and parents notice. A vague report does not just fail to inform; it actively signals that the academy is not paying close attention to this specific child, which is the exact opposite of what a report exists to prove. Vague reports are arguably worse than no reports, because silence at least does not leave evidence of inattention.

The academies that keep reporting alive past month three all make the same structural move: they separate evidence collection from judgment. Attendance flows from the register that is already taken every class. Games and rating movement flow from the student's synced Lichess or Chess.com account rather than from anyone's memory [2]. With the evidence assembling itself, the coach's job shrinks to the part only a coach can do: three or four sentences of judgment and one next step. That is a five-minute task per student, and five-minute tasks survive. The rest of this guide is about building that separation, and our pillar on running a chess academy covers where reporting sits among the other operating systems an academy needs.

Every number in a progress report should come from a system of record, never from memory or estimation. There are three systems that matter: the attendance register for classes, the synced game platform for games and ratings, and the engine for move quality. If a number cannot be traced to one of those three, it does not belong in a document a parent will keep.

Game data is the easiest to automate and the most commonly fudged by hand. Lichess publishes a full API that exposes a player's games, rating history, and performance statistics [2], which means a student's month of chess can be collected programmatically rather than transcribed from screenshots. Chess.com game history can likewise be imported for students who play there. Once games are synced, move-quality claims should come from Stockfish, the open-source engine that produces deterministic, reproducible evaluations [4]. The difference matters when a parent asks why the report says their child blundered twice: an engine evaluation is checkable, an impression is not.

Attendance is the number parents silently audit. A parent knows exactly how many classes their child attended because they did the dropping off, so an attendance figure that is wrong by even one class undermines every other number in the report. This is why attendance must flow from the register taken in class, on the day, and never be reconstructed at report time. Our guide to attendance tracking, linked from the academy operations pillar, covers the register habit itself; for reporting purposes the rule is simply that report numbers are read from records, not written into them.

• Attendance: read from the class register, recorded on the day of each class.
• Games played and rating movement: read from the synced Lichess or Chess.com account [2].
• Accuracy, blunders, and key moments: read from Stockfish analysis of those synced games [4].
• Fees and receipts: read from payment records, and kept out of progress reports entirely so money and progress stay separate conversations.
• Coach narrative: the only part written fresh, because judgment is the one thing no system of record can produce.

Talk about ratings as a noisy estimate over months, never as a score that should rise every week. The Elo system, which underpins chess ratings everywhere, measures a player's strength relative to the opponents they actually played, and the published number is an estimate that updates with every result [1]. A child can be improving while their rating dips for a month, and a child can plateau while their rating drifts up. Reports that treat every dip as a crisis and every rise as proof of teaching quality train parents to do the same, and that trade always comes due.

Be explicit about which rating you are reporting. A Lichess rapid rating, a Chess.com blitz rating, and a FIDE classical rating are three different numbers on three different scales, and they are not interchangeable. FIDE publishes its official rating list monthly [5], so an over-the-board rating moves at most twelve times a year no matter how much a student plays, while an online rating can move every evening. A report that says the rating went up without naming the pool and time control is technically true and practically misleading. Our spoke on showing parents rating progress goes deeper on charts and framing that make this legible.

The honest framing also protects the academy. The fastest way to lose a family is to let them believe the academy controls the rating. Ratings depend on opponent pools, rating deflation in junior sections, how often a child plays, and plain variance [1]. What the academy controls is the quality of training and the evidence of skill growth: cleaner endgame technique, fewer one-move blunders in synced games, better time management. Lead reports with those, and let the rating be the lagging indicator it actually is.

Monthly is the right cadence for almost every academy, and ownership should be split: the system assembles evidence, the coach writes judgment, and one named person owns the send. Weekly reports drown parents and burn coaches. Quarterly reports leave families silent for so long that the report arrives as a stranger. Monthly is frequent enough to show a story and rare enough that each edition is news.

Ownership is where most plans go vague, so make it explicit. The attendance register and game sync run continuously and belong to the system. The narrative and next step belong to the coach who actually teaches the student, never to an admin paraphrasing second-hand. The send belongs to one named person with a date: in the demo academy, reports for Batch B2 go out on the first Saturday of the month, and the owner checks one list to see which reports are approved and which still wait. A report with no owner and no date is a report that will not ship.

Delivery channel matters as much as cadence. Most academies in India deliver reports where parents already are, which is WhatsApp; WhatsApp's business tooling exists precisely because customers expect businesses to reach them there [3]. The trade-offs between a group chat, one-to-one threads, and a proper guardian portal are real, and we compare them honestly in our guardian portal versus WhatsApp group piece, with the operational details in our guide to WhatsApp for chess academies. Whatever the channel, the rule from earlier still applies: the report should invite a one-line reply, and someone must be assigned to answer it.

The workflow has four stages: the system gathers evidence, an AI drafts the narrative around injected numbers, a coach edits and approves in a queue, and only approved reports are delivered to families. The order is the point. The model never invents a number, because attendance, games, and ratings are injected from the systems of record described above, and a named human stands between every draft and every parent.

Drafting is where the month-two problem finally breaks. A language model given a student's verified month, twelve classes attended out of twelve scheduled, eighteen rapid games synced, two engine-flagged blunders in the same rook endgame pattern, can produce a competent first draft of the narrative in seconds. The coach's task changes from composing fifteen reports to reading fifteen drafts, correcting tone or emphasis on two or three, and approving the batch. Judgment stays human; only the blank page is gone. We cover the drafting mechanics in detail in our piece on AI-drafted progress reports.

The approval queue is the guardrail that makes the speed safe. Every draft waits in a single list showing the report, the records it was built from, and two actions: edit or approve. Nothing reaches a parent without a named human pressing approve, which keeps the coach as the author of record on every message the academy sends. Why that gate is non-negotiable, including what can go wrong without it, is the subject of our article on the AI approval queue.

Evaluate any tool that claims this workflow against three questions. Can a draft reach a parent without a named human approving it? Are the numbers injected from records, or does the model write them? Is there a trail showing who approved each report and when? The integrations behind the evidence layer, game sync from Lichess and Chess.com and attendance from the register, are what separate a reporting system from a text generator [2], and the AI layer should be judged on its guardrails before its prose.