One client, one article, one address: the same data everywhere.

A client calls about an invoice, and the salesperson finds them three times in the CRM, with three different addresses. Accounting sends the reminder to the old address. In the online store, the same article has a different weight than in the warehouse system, so the shipping costs are wrong. Nobody did this on purpose. It crept in over ten years, and now someone spends an afternoon every week working out which version is the right one.

Why your master data drifts apart

Master data is the information that does not change every day: clients, suppliers, articles, employees, locations. It lives in every system you use, and every system believes its own version is the right one. A new colleague enters a client again because the search did not find it. A supplier moves, and the address is updated in two of the four packages. An article number gets a hyphen in purchasing that is missing in sales.

The result is not one big error but a thousand small ones. They only cost you money when you start calculating with them. A report that adds up revenue per client counts Janssen BV and Jansen B.V. separately. An integration that passes on orders gets stuck on an article that exists in one system and not in the other. That is when it turns out the data was the problem, not the software.

Measure first, then clean

We do not start by tidying up. We start by counting. On an export of your clients, articles, and suppliers, we measure how big the problem is: how many records are probably duplicates, which fields are structurally empty, how many addresses do not exist according to the postal code table, how many articles have no unit or weight. You get a list of counts, not an opinion. Often 80 percent of the mess turns out to sit in two or three fields.

Then we clean up. We use AI to find duplicates, because a person sees that Jansen, Janssen, and JANSSEN B.V. belong together, but a simple search does not. For the decisions, we do the reverse: the software puts the doubtful cases side by side, and someone on your side who knows the clients chooses. We log everything we merge or change, so a wrong choice can be undone. We process personal data on a server in the EU, in line with the GDPR, and the cleaned data is yours afterward.

One source, and the rest follows

Cleanup alone helps for a few months. Then the first client appears twice in the CRM again. What lasts is agreeing, for each data item, which system is the source. A client's address comes from accounting. An article's weight comes from the warehouse system. The other packages may show that data, but not change it themselves. That rule is simple to write down and hard to keep up without technology behind it.

That is why we connect the source to the rest. If the address changes in the source system, it moves automatically to the online store and the CRM. We often do this work together with Data pipelines. If you are switching to new software at the same time, we combine it with Data migration, because that is the only moment when you do not have to carry the dirt along. Within ten business days, you have a working prototype of one data flow, such as the client address, so you see how it feels before we tackle the rest.

Keeping it clean is the real work

The data is clean on the day we finish. The question is what it looks like a year from now. That is why we build checks at the moment of entry: a client cannot be saved without a Chamber of Commerce (KvK) number, an article cannot be saved without a unit, and whoever creates a new client first sees the three that already look like it. That is less work than it sounds, and it helps more than one big annual cleanup.

Whatever slips through goes into a monthly report. This many new duplicates, this many empty fields, and who entered them, so you can fix the cause instead of the consequence. We record in your Data governance who may change which data item and who owns it. If you like, the figures from this report land directly in your Data and reporting. That way it is not a project done once, but a habit the company keeps in its own hands.