A Data Role Is a Centre of Gravity, Not a Box
A practitioner view of how Data Engineering, Analytics Engineering, BI, Analysis, Architecture and Governance overlap across the data work stack.
Articles
Short articles on why business numbers do not match, why teams pay an Invisible Data Tax, and what to inspect first.
Recent additions
Newly published notes appear here first so readers and crawlers can reach fresh implementation material without waiting for a topic page to be reorganised.
A practitioner view of how Data Engineering, Analytics Engineering, BI, Analysis, Architecture and Governance overlap across the data work stack.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
Why durable AI memory needs structured contracts, provenance, tests, and modeled context before agents can safely answer business questions from notes.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
A practical way to use AI for analytics engineering while keeping grain, metric definitions, reconciliation, access, and repository health under human control.
Topic Hub
Articles for leaders searching by the symptom: revenue disputes, finance vs sales mismatch, dashboard disputes, and spreadsheet reconciliation.
Why spreadsheet logic often contains the business meaning AI and dashboards need, and how to turn that hidden semantic layer into Reporting Trust.
What changing board pack numbers reveal about definitions, ownership, timing, caveats, and reporting trust.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
How to diagnose revenue reporting disputes when sales, finance, operations, and dashboards all show different numbers.
What to inspect when a revenue dashboard, sales report, CRM view, spreadsheet, and finance pack show different numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Why month-end reporting slows down, what the delay reveals about reporting trust, and what to inspect before asking for faster dashboards.
What source-to-report lineage means, where it fits inside the Data Value Chain, and what leaders should inspect when business numbers do not match.
The hidden business cost of checking, reconciling, rebuilding, and explaining reports that people do not fully trust.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
Topic Hub
Core explanations for leaders trying to understand why trusted reporting breaks.
A practical walkthrough for exploring semi-structured event data, normalising JSON fields, and building reusable SQL models before answering business questions.
A practical way to use AI for analytics engineering while keeping grain, metric definitions, reconciliation, access, and repository health under human control.
Why durable AI memory needs structured contracts, provenance, tests, and modeled context before agents can safely answer business questions from notes.
A practical way to inspect one contested KPI when teams disagree about the number, its definition, source, owner, caveats, and safe use.
How dbt models, Looker semantic logic, metric ownership, lineage, and business definitions work together to create Reporting Trust.
Where AI can help from Capture to Realise, why the Semantic Layer Gate matters, and how to avoid accelerating weak reporting logic.
How Reporting Trust and semantic layer trust differ, where they overlap, and why the semantic layer is a control point rather than the buyer-facing outcome.
Why AI tools need controlled metric meaning, caveats, ownership, and safe-use rules before they can reliably answer business questions.
Why spreadsheet logic often contains the business meaning AI and dashboards need, and how to turn that hidden semantic layer into Reporting Trust.
What changing board pack numbers reveal about definitions, ownership, timing, caveats, and reporting trust.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
Why AI makes reporting trust more important, not less, and what teams should fix before automating decisions from messy numbers.
How to diagnose revenue reporting disputes when sales, finance, operations, and dashboards all show different numbers.
What to inspect when a revenue dashboard, sales report, CRM view, spreadsheet, and finance pack show different numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Why month-end reporting slows down, what the delay reveals about reporting trust, and what to inspect before asking for faster dashboards.
What source-to-report lineage means, where it fits inside the Data Value Chain, and what leaders should inspect when business numbers do not match.
The hidden business cost of checking, reconciling, rebuilding, and explaining reports that people do not fully trust.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
Topic Hub
How teams reduce decision drag, reconcile numbers, and rebuild confidence in priority metrics.
A practical walkthrough for exploring semi-structured event data, normalising JSON fields, and building reusable SQL models before answering business questions.
A practical way to use AI for analytics engineering while keeping grain, metric definitions, reconciliation, access, and repository health under human control.
Why durable AI memory needs structured contracts, provenance, tests, and modeled context before agents can safely answer business questions from notes.
A practical way to inspect one contested KPI when teams disagree about the number, its definition, source, owner, caveats, and safe use.
How dbt models, Looker semantic logic, metric ownership, lineage, and business definitions work together to create Reporting Trust.
Where AI can help from Capture to Realise, why the Semantic Layer Gate matters, and how to avoid accelerating weak reporting logic.
How Reporting Trust and semantic layer trust differ, where they overlap, and why the semantic layer is a control point rather than the buyer-facing outcome.
Why AI tools need controlled metric meaning, caveats, ownership, and safe-use rules before they can reliably answer business questions.
Why spreadsheet logic often contains the business meaning AI and dashboards need, and how to turn that hidden semantic layer into Reporting Trust.
What changing board pack numbers reveal about definitions, ownership, timing, caveats, and reporting trust.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
Why AI makes reporting trust more important, not less, and what teams should fix before automating decisions from messy numbers.
How to diagnose revenue reporting disputes when sales, finance, operations, and dashboards all show different numbers.
What to inspect when a revenue dashboard, sales report, CRM view, spreadsheet, and finance pack show different numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Why month-end reporting slows down, what the delay reveals about reporting trust, and what to inspect before asking for faster dashboards.
What source-to-report lineage means, where it fits inside the Data Value Chain, and what leaders should inspect when business numbers do not match.
The hidden business cost of checking, reconciling, rebuilding, and explaining reports that people do not fully trust.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
Topic Hub
What to fix before AI, automation, or generated summaries scale weak reporting logic.
A practical way to use AI for analytics engineering while keeping grain, metric definitions, reconciliation, access, and repository health under human control.
Why durable AI memory needs structured contracts, provenance, tests, and modeled context before agents can safely answer business questions from notes.
How dbt models, Looker semantic logic, metric ownership, lineage, and business definitions work together to create Reporting Trust.
Where AI can help from Capture to Realise, why the Semantic Layer Gate matters, and how to avoid accelerating weak reporting logic.
How Reporting Trust and semantic layer trust differ, where they overlap, and why the semantic layer is a control point rather than the buyer-facing outcome.
Why AI tools need controlled metric meaning, caveats, ownership, and safe-use rules before they can reliably answer business questions.
Why AI makes reporting trust more important, not less, and what teams should fix before automating decisions from messy numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Topic Hub
Practical inspection points for definitions, source paths, ownership, and reporting quality.
A practical way to inspect one contested KPI when teams disagree about the number, its definition, source, owner, caveats, and safe use.
How Reporting Trust and semantic layer trust differ, where they overlap, and why the semantic layer is a control point rather than the buyer-facing outcome.
Why AI tools need controlled metric meaning, caveats, ownership, and safe-use rules before they can reliably answer business questions.
Why spreadsheet logic often contains the business meaning AI and dashboards need, and how to turn that hidden semantic layer into Reporting Trust.
What changing board pack numbers reveal about definitions, ownership, timing, caveats, and reporting trust.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
Why AI makes reporting trust more important, not less, and what teams should fix before automating decisions from messy numbers.
How to diagnose revenue reporting disputes when sales, finance, operations, and dashboards all show different numbers.
What to inspect when a revenue dashboard, sales report, CRM view, spreadsheet, and finance pack show different numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Why month-end reporting slows down, what the delay reveals about reporting trust, and what to inspect before asking for faster dashboards.
What source-to-report lineage means, where it fits inside the Data Value Chain, and what leaders should inspect when business numbers do not match.
The hidden business cost of checking, reconciling, rebuilding, and explaining reports that people do not fully trust.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
A practitioner view of how Data Engineering, Analytics Engineering, BI, Analysis, Architecture and Governance overlap across the data work stack.
A practical walkthrough for exploring semi-structured event data, normalising JSON fields, and building reusable SQL models before answering business questions.
A practical way to use AI for analytics engineering while keeping grain, metric definitions, reconciliation, access, and repository health under human control.
Why durable AI memory needs structured contracts, provenance, tests, and modeled context before agents can safely answer business questions from notes.
A practical way to inspect one contested KPI when teams disagree about the number, its definition, source, owner, caveats, and safe use.
How dbt models, Looker semantic logic, metric ownership, lineage, and business definitions work together to create Reporting Trust.
Where AI can help from Capture to Realise, why the Semantic Layer Gate matters, and how to avoid accelerating weak reporting logic.
How Reporting Trust and semantic layer trust differ, where they overlap, and why the semantic layer is a control point rather than the buyer-facing outcome.
Why AI tools need controlled metric meaning, caveats, ownership, and safe-use rules before they can reliably answer business questions.
Why spreadsheet logic often contains the business meaning AI and dashboards need, and how to turn that hidden semantic layer into Reporting Trust.
What changing board pack numbers reveal about definitions, ownership, timing, caveats, and reporting trust.
Why dashboards and spreadsheets show different numbers, and what leaders should inspect before rebuilding reports.
A practical diagnostic checklist for leaders when dashboards, spreadsheets, and finance packs show different numbers.
Why AI makes reporting trust more important, not less, and what teams should fix before automating decisions from messy numbers.
How to diagnose revenue reporting disputes when sales, finance, operations, and dashboards all show different numbers.
What to inspect when a revenue dashboard, sales report, CRM view, spreadsheet, and finance pack show different numbers.
A leader-friendly guide to the KPI definition decisions that make dashboards easier to trust, reconcile, and use.
Why month-end reporting slows down, what the delay reveals about reporting trust, and what to inspect before asking for faster dashboards.
What source-to-report lineage means, where it fits inside the Data Value Chain, and what leaders should inspect when business numbers do not match.
The hidden business cost of checking, reconciling, rebuilding, and explaining reports that people do not fully trust.
Why dashboards, spreadsheets, and finance packs drift apart, and what to inspect first when leaders stop trusting the numbers.
Free chapter
Download the opening chapter of Your Numbers Don't Match and start spotting where reporting trust breaks down.