Data modernization starts with where it comes from
Data programmes rarely fail on the platform. They fail on the sources: figures that say something different in four places, integrations run by hand, and nobody who owns a definition.
Four problems no platform solves
They all surface the moment you put two sources side by side, and none of them disappears by moving somewhere else.
- Same term, different meaningWhat counts as a "customer" differs between sales, invoicing and delivery. As long as nobody chooses, every report counts something different and loses out to a private list in a spreadsheet.
- Integrations someone runs by handAn export on Friday, an import on Monday, a colleague who knows. That works until the colleague leaves, and then it is not broken once but structurally.
- Sources nobody maintains any moreA system that only supplies data and is otherwise left alone. It falls over on the day the certificate expires or the server runs out of support.
- No owner per figureWithout someone who decides what comes from where, every discrepancy becomes a debate instead of a correction. This is the only one of the four that is not technical, and the only one you have to solve first.
Where we add value
The layer under a data programme, and the people to carry it. Not the programme itself.
The infrastructure underneath
Storage, network, rights, backup and the environment a platform runs in — on your own equipment, in the cloud or both. That is the part that should be ready before anything is built.
Access and separation
Who can reach which data and how that is recorded. In a data programme, data from separated systems comes together, and then access control is not a detail but the core of the risk.
What we do not do
Design data models, build reporting or set up an analytics platform. We will say what the infrastructure underneath can carry and what changing that would cost.
Further reading
What this connects to.
Questions we get about this
The ones that come up most, answered briefly.
Why do data programmes stall on the sources?
Because the same term means something different in different systems and nobody chooses. Until that is resolved, every report counts something different and loses out to a private list in a spreadsheet.
What do you actually do in this?
The layer underneath: storage, network, rights, backup and the environment a platform runs in, plus access control on data coming together from separated systems. We do not build data models or reporting.
Get the layer underneath in order
Tell us what sources exist and where it has to go. We say what the infrastructure can carry today and what has to happen before anything is built.
Practical IT knowledge in your inbox
New guides on management, security and the workplace, written by the people doing the work. No sales talk, and you can unsubscribe in one click.