--- # /blocks/reporting/ - internal management reporting as a block (2026-07-24). # Split out from the old performance page, which conflated two things: the # recurring management pack (an artefact on a clock - this page) and the # operator's cockpit (an always-on view - the substrate, see performance.md). # Scope is internal management information; regulatory and statutory returns # belong to the obligations block. title: 'The reporting block' slug: 'reporting' object: 'Each management reporting cycle - the pack the exec and the board actually run on.' pain: 'The report takes weeks to build, and by the time it arrives the numbers have been shaped by everyone who touched it.' description: 'Internal management reporting run as one loop - the pack assembled from the live record, every figure traceable to its source.' metaTitle: 'The reporting block: a pack you can stand behind' related: - 'performance' - 'product' order: 3 noindex: false # launched 2026-07-24 --- # The reporting block **Internal management reporting, run as one loop** - the pack assembled from the live record, every figure carrying its source, and nothing distributed that cannot be stood behind. --- ## The pain *"The report takes weeks to build and distribute - and by the time it arrives, the numbers have been shaped by everyone who touched it. Nobody wants to be criticised."* The management pack is manufactured, not read. Each department extracts its own numbers from its own system, summarises them in the way that reflects best on the department, and passes them up. Every hand-off adds delay and subtracts candour. By the time the pack reaches the people who decide, it describes a business that existed weeks ago, filtered through the incentives of everyone in the chain - and a significant share of mid-level effort exists purely to move those numbers around. Then the first question in the meeting is "where did this figure come from?", and nobody can answer it in the room. In professional services this is the monthly partner pack; in distribution, the trading review; in regulated firms, the board MI. ## What the block is The object is the management reporting cycle - one continuous record per cycle, from period close through assembly, challenge and sign-off to the questions the pack provokes. Everything it draws on feeds that record: the source figures and where they came from, contributions and who made them, adjustments and their reasons, metric definitions and when they changed, prior periods and plan. A model reads it permanently. Your finance system, CRM and ledgers stay exactly where they are - the block reads from them, it does not replace them. ## The loop in action [MOMENT: styled quote panel] The period closes and the pack assembles itself: revenue by line, WIP ageing, pipeline movement, each figure carrying its source and its variance against plan and prior period, with the narrative drafted. *"Here is the pack; every number traces to the record it came from."* The extracting, chasing and reconciling that took a fortnight of mid-level effort is simply not done by people any more. - *the dial: draft* [MOMENT: styled quote panel] The catch. A figure was adjusted by hand after submission, and no reason was recorded. The block holds distribution and puts both numbers in front of the named owner: *"this was submitted at one value and sourced at another; the pack does not go out until someone owns the difference."* The distortion that used to survive three layers of reporting is caught before the board reads it, not after a decision has been taken on it. - *the dial: recommend, with the governance holding* [MOMENT: styled quote panel] A challenge comes back on a distributed pack - why is this line down? The answer is assembled by walking from the number to the underlying records and written with its evidence attached, in one step rather than one meeting and three days. The definition drift behind it - this quarter's measure is not last quarter's - is flagged with a fix proposed, so the trend line means something. - *the dial: draft* Each output states what it saw, why it matters, what it recommends, and how confident it is - every fact linked back to its source, in an order an operator can read and a board or an auditor can examine. ## You set the dial Every capability has four positions: **observe** - it reads silently and measures its own accuracy; **recommend** - it surfaces the case and a proposed action to a named owner; **draft** - it prepares the action for a human signature; **act** - it executes within agreed bounds and logs everything. Everything starts at observe. Promotion is earned on the measured record and reversed instantly if it doesn't hold. Nothing is distributed without the sign-off your policy requires, and who sees which cut of the pack stays a setting you control - permanently, if that is your policy. The mechanics are the immune gate described in [our framework](/ai-native-company/). ## How it lands It starts narrow: one pack - the monthly management accounts, or the exec scorecard - the block at observe, assembling in parallel with the report your team builds today so you can judge it against your own known numbers. Nothing is replaced, nothing is migrated - the first figures flow from the systems you already run. From there it is a bounded engagement - blueprint, build, handover - and your team owns the result outright: the signal map, the judgement rules, the dial settings, the working loop. ## The backtest Take three past packs: one that landed too late to act on, one whose number was challenged and could not be traced, and one that surprised the board. What did your own systems know, and when? That is the first question the diagnostic answers. Take [the ten-minute diagnostic](https://benchmark.somai.studio), or write to [hello@somai.studio](mailto:hello@somai.studio?subject=The%20reporting%20block). --- Runs on [our framework](/ai-native-company/) ยท related reading: [Most AI spend never reaches the P&L](/thinking/most-ai-spend-never-reaches-the-pnl/)