From 40 metrics nobody trusts to a KPI framework that drives decisions
Defining a North Star and KPI tree for a marketplace scale-up - and wiring definitions into code so they stay true.
Read → AI & AutomationBuilding AI data agents your CFO can trust
The model never generates a number: verifiability principles for analytics agents that survive real scrutiny.
Read → Working TogetherWhy a fractional data expert beats your first data hire
The honest math on senior fractional expertise vs a full-time hire - and when a fractional is the wrong choice.
Read → ExperimentationWhy smart teams still guess wrong - the case for A/B testing
1.2 million charity envelopes, two clever nudges that lost money, and how to build an experimentation practice that pays.
Read → Analytics EngineeringRescuing a 2-year-old dbt project
The audit we run before changing a single model - and the hardening checklist that puts a project ahead of 90% of dbt setups.
Read → AI & AutomationAI data analyst: why context beats a bigger model
The "start ticket" agent workflow we use on client work - and why structure, not model upgrades, took accuracy from 26% to 94%.
Read → Growth AnalyticsCustomer retention analysis: the four fingerprints of a retention drop
Product break, calendar shock, market drift or mix shift? Each leaves a different fingerprint on the cohort grid.
Read → Probabilistic ModellingCustomer lifetime value prediction: every customer flips two coins
Answering "what is a customer worth?" for a non-contractual e-commerce - with prediction error most cohort models can't touch.
Read →