business-intelligence
Builds reporting and self-serve analytics that people actually use — metric trees, dashboard design, distribution, and the discipline that stops dashboards proliferating. Use this to build a dashboard or report, design a metrics framework, set up self-serve analytics, decide what to measure, or diagnose why reporting exists but nobody uses it or trusts it.
git clone --depth 1 https://github.com/cbrock84/headcount /tmp/business-intelligence && cp -r /tmp/business-intelligence/plugins/data-analytics/skills/business-intelligence ~/.claude/skills/business-intelligenceSKILL.md
# Business intelligence Most organizations have too many dashboards and too little insight. The two are related: when everything is measured, nothing is watched. ## Start from the decision Every report answers one question for one audience who can act on it. Before building, name the decision it informs and what a viewer would do differently based on it. If nothing would change, do not build it. That single filter removes most dashboard requests, and the ones surviving it get used. ## Metric trees Structure metrics as a tree, not a list. One primary outcome at the top, decomposed into the drivers that mathematically produce it, each decomposed again. Revenue = customers × average value. Customers = new + retained. New = traffic × conversion. And so on. This does two things a metric list cannot: when the top number moves, you can walk down to find *where*; and it makes clear which metrics are levers and which are outcomes. Teams should be measured on levers they control, not on outcomes they influence. ## Dashboard design - **One screen, one question.** Scrolling dashboards are several dashboards that were not separated. - **Lead with the answer** — the primary number, its comparison, and whether that is good. A number with no comparison is not information. - **Comparison always**: prior period, target, or cohort. Choose deliberately, because each tells a different story. - **Say what "good" is.** A viewer who cannot tell whether 4.2% is good will not act. - **Annotate the anomalies.** The spike everyone asks about should carry its explanation, or you will explain it every month. - **Cut the rest.** Charts nobody uses cost attention on every visit and make the useful ones harder to find. ## Self-serve Self-serve works when the semantic layer is trustworthy and the questions are anticipated. It fails when people are handed raw tables and left to define metrics themselves — that produces confident wrong answers, which is worse than a queue. Give governed metrics, curated datasets, and templates for common questions. Keep the raw layer for analysts. ## Trust Reporting nobody trusts is not used, and trust is lost far faster than it is rebuilt. Protect it by showing freshness on every dashboard, surfacing failures rather than serving stale data silently, and reconciling against the system of record for anything financial. When a number is wrong, say so prominently and fast. Quietly correcting it is how a team learns to check every figure by hand. ## Maintenance Dashboards accumulate. Review usage periodically and retire what nobody opens — with a notice period, since the one person using it may be using it for something important. ## Tooling BI: Power BI, Looker, Tableau, Metabase, Omni, Hex, and similar. Metric definitions belong in a semantic layer — dbt's, Looker's LookML, Cube — rather than in each dashboard's SQL, or the same metric will disagree with itself across two tabs. Spreadsheets remain the most-used BI tool in every organization. Plan for the export rather than pretending it will not happen. ## Never - Build a dashboard nobody has a decision for. Start from the decision. - Ship a metric with two definitions live at the same time. - Leave a dashboard published with no owner. Unowned dashboards get trusted, then get wrong. - Show a number without the denominator and the window it covers.
Concentrates marketing and sales effort on a named set of accounts rather than on volume — qualifying whether the model fits your economics at all, building the account list and the buying group inside each, tiering effort against account value, coordinating so the account experiences one campaign rather than several, and measuring account progression instead of leads. Use this to decide whether to run an account-based program, build one, or work out why an existing one produces activity and no pipeline.
Designs orchestrator-and-subagent hierarchies for a repository — splitting agents by exclusive write surface, pairing every producer with an independent auditor, and enforcing the split with a script that runs in CI. Use this whenever the user wants to set up, expand, audit, or fix a multi-agent or subagent structure for a codebase; asks how to divide work between agents; wants agent charters, roles, or a surface map written; or is hitting agents that collide on the same files, review their own work, or drift from their remit. Also use when sizing a roster or deciding whether a new agent is justified.
Governs models and AI systems in production — intended use, evaluation, monitoring, human oversight, documentation, and the decision to deploy or retire. Use this before deploying a model or AI feature, when defining evaluation criteria, when a model's behavior has drifted, when assessing AI risk or regulatory exposure, or when deciding whether an AI system is fit for a consequential decision.
Produces executive-level research — market sizing, competitor mapping, trend analysis, and strategic intelligence — grounded in cited sources with the confidence in each claim made explicit. Use this to analyze a market or industry, map competitors, evaluate a market-entry or build-versus-buy decision, produce a research brief, or assemble evidence for a decision. Also use when comparing options that need a structured, evidence-based verdict rather than an opinion.
Optimizes for AI assistants and AI-generated answers — being retrievable, being cited, and being represented accurately when a model answers on your behalf. Use this when traffic is shifting from links to AI answers, when a brand is misrepresented or absent in AI responses, when planning content for retrieval rather than ranking, or when deciding how AI search changes an existing SEO program.
Improves visibility and conversion in the App Store and Google Play — metadata, keywords, screenshots, ratings, and the listing experience that turns an impression into an install. Use this to audit or optimize an app listing, plan a launch listing, diagnose poor install conversion, or improve store search visibility.
Protects and restores data — backup coverage and scope, retention, immutability against ransomware, and proving restores actually work. Use this to design a backup regime, verify restores, plan retention, protect backups from ransomware, or recover from data loss.
Runs the planning cycle — annual budget, rolling forecast, consolidation of business unit inputs, and the variance analysis that explains actuals against plan. Use this to build or challenge a budget, run a forecast cycle, consolidate departmental submissions, explain why results diverged from plan, or improve forecast accuracy.