TMI / GLOSSARY
DATA FOUNDATION
Definition. A data foundation is the work of aggregating, connecting, cleaning, digitising and structuring what a company already knows, until its records can carry a decision.
WHAT IT
MEANS.
It is the least exciting stage of any technology programme and the one whose absence kills the rest. Reporting built on records that disagree produces confident wrong answers. Automation built on them industrialises the error. Anything called AI built on them is a very expensive way to be misled.
The work itself is concrete: bring the records into one place, make the systems speak to each other, clean what is dirty, digitise what is still on paper, capture what only exists in someone's head, and agree what the source of truth is when two systems disagree.
WHAT IT
IS NOT.
A data warehouse
A warehouse is one possible component. A foundation is the whole job, and in most established companies the expensive part is the paper and the people, not the storage.
A one-off cleanup
A cleanup degrades from the day it finishes unless something upstream changed. The foundation includes deciding how records stay right.
Every term in this glossary is one TMI uses in client work, which is the only reason it is here. The rest of the glossary →
SEE IT IN YOUR OWN COMPANY