The Dashboard Design Checklist
27 yes-or-no checks to review a dashboard before anyone acts on it. Six groups cover design quality, and three cover whether the dashboard answers the question it was built for.
My writings on data strategy, AI, governance, and how to turn data into a strategic asset that informs every decision.
27 yes-or-no checks to review a dashboard before anyone acts on it. Six groups cover design quality, and three cover whether the dashboard answers the question it was built for.
For twelve months, my LinkedIn feed was full of posts saying dashboards did not matter any more, because a large language model can answer any question in text. Then Claude Dashboards came out, and within a week the same feed was full of people celebrating dashboards again.
Corporate real estate keeps selling its capital programme. Its real value is the property hierarchy, the anchor the rest of the organisation and its AI can be measured against.
Part 2 of 2. Corporate real estate has no data foundation for the AI era. Microsoft may have already built one, almost by accident. Where the evidence for these decade-long bets already sits.
Part 1 of 2. Everyone is asking how much office space to cut. Almost nobody is asking whether voice-first AI collaboration means we need more, not less.
Six personal projects, six rungs of a ladder. Vibe coding gets you to the bottom fast. Getting further up is a different game entirely.
Five habits that get more out of AI assistants like ChatGPT, Claude and Copilot. None of them are about the model.
Five iterative steps from raw data to certified, boardroom-ready insight. The canonical reference.
Six principles for how a data and analytics team should actually work. Design out, ship in.
Why Large Language Models depend on semantically conformed data — and why Step 2 is the AI-readiness gate.
Step 1 deep-dive. The deliberate refusal to start until the question is properly defined.
Step 5 deep-dive. Explicit verdicts, visible marks, no soft sign-off.