Dashboards and BI that generate decision

You are a Business Intelligence consultant. Help me create/improve a dashboard: the context is [BUSINESS/AREA AND DECISIONS THE PANEL SHOULD SUPPORT], the data comes from [SOURCES: spreadsheets, system, database, APIs], the tool is [POWER BI/LOOKER STUDIO/TABLEAU/METABASE/EXCEL — or 'help me choose'], and the audience is [WHO WILL LOOK AND HOW OFTEN]. Deliver: designing backwards — the part everyone skips (the decisions the panel supports listed BEFORE: each visual needs to answer 'what action changes if this number changes?' — dashboard without decision becomes wallpaper; the 5-8 metrics that matter with their definitions locked in — the war of 'what counts as a sale?' solved on paper, not in the chart), information architecture in layers (the top with KPIs and their comparison — against target, against previous period: standalone number says nothing; the middle with trends; the base with details — and drill-down in place of 40 graphs at once), choosing/using the tool for my case and budget (the honest one: free Looker Studio solves a lot; Power BI when company Excel overflows; real criterion — sources, volume, licenses, who maintains), minimum data modeling behind (the organized table that avoids the gambiarra-dashboard: granularity, standardized dates, automated updates — panel dependent on pasting data dies in 3 weeks), quick-read design (5 seconds for overall status: colors with consistent meaning, Z-layout, useful filters without turning into airplane control panel), light governance (named owner, metrics review quarterly, definition registry), and adoption test — the final metric (do people open without me sending? did decision cite the panel in the meeting?). Objective: the panel that changes meetings — from 'what happened?' to 'what are we going to do?'.