Master data workflow: how the record approval flow works

Master data workflow is the path a record travels between the creation request and the moment it becomes an official record in the ERP, passing through validation, enrichment, and approval along the way. I spent years implementing this flow inside SAP, as a certified Workflow consultant, and I learned one thing early: no system approves a record on its own. People approve, following a rule the company defined beforehand.

That is the point most companies get wrong. They buy an MDM tool, roll out the new ERP, but leave the approval flow the old way: an email to the boss, a shared spreadsheet, a "send it on WhatsApp and I'll release it". The result is duplicate materials, suppliers without active approval, customers with a CNPJ (Brazil's company registration number) that no longer even exists at the Receita Federal (Brazil's federal tax authority).

What master data workflow is

Master data workflow is the structured process that takes a request to register, create, or change a material, customer, or supplier from origin to publication in the system, with defined steps for who requests, who validates, who enriches, and who approves.

It is the flow that decides whether that screw will be registered for the third time or whether the system will block it because an identical one already exists. It is what decides whether a supplier enters the database with expired documentation or is held back until it is regularized.

In the SAP Master Data Governance projects I worked on, the workflow is configured with routing rules: an MRO material request goes to one approver, a raw material request goes to another, an international supplier goes through an extra compliance step. Without that structure, the master record becomes land with no owner, and land with no owner is where duplication and error settle in.

How the flow works in practice

The master data approval flow tends to follow a similar sequence in any mid-sized or large company, regardless of the ERP.

The request is born with whoever needs the data: a buyer asking for a new supplier, a salesperson asking for a customer, a maintenance engineer asking for a material, filling in a form in the master data portal or directly in the ERP.

Then comes automatic validation. The MDM system checks whether a similar record already exists, whether the CNPJ is valid, whether the registration status at the Receita Federal is regular, and whether the mandatory fields were filled in, preventing the request from reaching the approver's desk incomplete.

Next comes enrichment. The data steward fills in what is missing: tax classification, material group, unit of measure, the supplier's bank details, delivery address, using the company's descriptive standardization of materials and taxonomy so that the new record speaks the same language as the rest of the database.

Then the request goes to approval. The data owner in the business area confirms that this material, customer, or supplier really should exist, with those characteristics. In larger companies, this happens in parallel: procurement approves the commercial side, tax approves the tax classification, quality approves the technical vetting.

Only after all approvals is the record published in the ERP, whether SAP, TOTVS, or Oracle, and it becomes valid for purchase orders, nota fiscal (Brazilian tax invoice), and inventory. Each step is logged, with date, owner, and decision, forming the audit trail that any compliance area will ask for sooner or later.

Who takes part in this flow

The master data approval flow crosses several areas, and each has a different role.

The requester sees the need for the record day to day, usually in procurement, sales, engineering, or maintenance. The data steward looks after data quality, fills in fields, applies the standardization rule, and sends the request back when information is missing. The data owner has the business authority to approve, because they understand the impact of that material or supplier on the operation. IT and governance maintain the flow design, adjust the routing rules, and look after the integration with the ERP.

In exception cases, such as a supplier that fits no standard rule, the data committee steps in to decide, with an agenda, an owner, and a deadline, instead of leaving the request sitting in a forgotten inbox.

Which tools automate the workflow

Market Master Data Management platforms offer a workflow engine as one of their core features, precisely because the manual process does not scale.

SAP Master Data Governance, native to SAP itself, configures approval flows by data domain, with business rules in BRFplus and direct integration with S/4HANA. Informatica MDM and Stibo Systems offer configurable workflow engines, with parallel and sequential approval, integrated with multiple ERPs. Reltio and Semarchy bring native cloud workflow, for companies with several source systems. Profisee and Ataccama, for their part, add a strong data quality component to the workflow, blocking at the door whatever does not meet the defined rules.

None of these tools replaces the business decision. They automate the routing, enforce deadlines, log the trail, and prevent the record from advancing without the right approval. The one who decides remains the process owner inside the company.

How to learn to design a master data workflow

The first step is to understand the business process that will become a flow, even before opening any tool: who requests, who needs to approve, what happens when information is missing, what counts as an exception. A tool without a designed process just becomes a nice system recording the mess faster.

From there, it is worth studying the DAMA-DMBOK, the reference guide to data management, which covers workflow and approval within the master data operations pillar. The CDMP certification, from DAMA itself, helps formalize that knowledge and is well regarded in the market. Those who already work with SAP can complement it with specific training in Master Data Governance.

But the part that teaches the most is taking part in a real project, designing the flow with the business areas before configuring any rule in the system. That is where you learn the difference between the flow the manual describes and the one the company actually uses.

Career opportunities

Mid-sized and large companies that run SAP, TOTVS, or Oracle, and that handle a high volume of material, customer, and supplier master data, need people who know how to design and maintain this workflow. This professional shows up under titles such as master data analyst, data steward, data governance analyst, or MDM consultant.

Demand grows alongside migration projects to new systems, such as S/4HANA, and the ever-increasing use of AI in companies, because no artificial intelligence agent makes good decisions on top of poorly approved records. Those who understand business process and translate it into workflow rules have room both at companies that are MDM customers and at consultancies and technology vendors.

Main questions about master data workflow

Is master data workflow the same thing as data governance? Governance is the set of policies, roles, and indicators that define how data should be handled. The workflow is the automated process that applies that governance in the day-to-day of the record.

Does every company need a formal approval workflow? Every company that has more than one person registering materials, customers, or suppliers already has a flow, formal or informal. What changes is whether it is documented, with an owner and a deadline, or whether it depends on the memory of whoever has been at the company longest.

How long does it take to implement this workflow? It varies with the number of data domains and areas involved, but an initial design, with the material or supplier flow mapped and configured in a tool, is usually ready in a few months when the business process is already clear.

Does the workflow slow down the pace of purchasing and sales? When well designed, no. The goal is to approve quickly what is correct and block early what is wrong, avoiding weeks of rework later, when the purchase order has already been issued to a supplier who should not have been approved.

Is it possible to automate the entire flow, without human approval? Part of the validation can be automatic, such as duplicate checking and CNPJ lookup. The decision to approve a new record remains a business one, because it involves financial, tax, and operational risk that no automatic rule covers on its own.

MDM Academy: where to go deeper on this with people who have implemented the flow

MDM Academy is 4MDG's data governance and master data school, created to train the professional who will build and maintain this kind of flow inside their own company. The courses cover governance, data quality, material master data, and preparation for certifications such as the CDMP, always with real project examples.

Those who want to learn to design an approval workflow without reinventing the wheel on every project will find a structured path there, with people who have already configured this process inside SAP, TOTVS, and other platforms. It's worth getting to know MDM Academy or joining the Data Club, the community that brings together people who live this routine of master data, procurement, and governance day to day.

Conclusion

Master data workflow is not extra bureaucracy. It is what keeps the record from becoming nobody's responsibility. Every company that grows goes through the moment when the master data no longer fits in a spreadsheet and a verbal agreement among colleagues. The difference between those who solve this well and those who keep putting out fires lies in the flow: who requests, who validates, who approves, and what happens when something falls outside the rule.

Thinking about your company's flow, who approves a new supplier today before it enters the system?