How do you manage data quality issues and conflicts in MDM, such as duplicates, inconsistencies, or errors?

Data Quality effort is the pre-requisite for building the Master Data. All sorts of Data Quality issues should be identified in that effort, that includes duplicates. In order to identify the Data Quality issues, data needs to be profiled with business rules applied on it. While duplicates can be handled as part of MDM solution, by doing "Match & Merge", the other issues should be handled out of the MDM box. Without addressing the Data Quality issues, the MDM solution will not yield efficient results.

You address data quality issues and conflicts in your data by having well defined data governance policies which clearly articulate the business rules that will be used to resolve these issues. For example, the ability to identify duplicate records is a function of having rules defined for what uniquely defines a given object/entity, and the same is true for resolving the duplicate itself. Without clearly defined policies, data stewards are left to use whatever rules they like to try to tackle these issues, which will lead to inaccurate and inconsistent approaches - the exact opposite of what you're trying to accomplish with MDM in the first place.

In MDM, managing data quality involves identifying issues, analyzing root causes, cleansing data, standardizing formats, establishing governance, appointing stewards, and continuous monitoring.

First of all, certain rules are self-evident and quite certainly a "must have" for MD. For example, the uniqueness (i.e. absence of duplicates). No need to go further otherwise your semantic model is not sound. Another example - domain definitions. You you create a column for e-mails, why on Earth you need to prioritise any validation rules? Grab a regex and implement the rule. For more complicated domain checks you may need some planning. Anyway, don't wait for problems to appear in order to "identify them" - treat them at the modelling phase.

Maintaining data quality is critical to make evidence-based decisions. to improve the data quality plan the data collection tool is simple, integrated, and standardization throughout implementation sites.

Data quality rules need to be established for all data, including master data. If you've established governance first, you've identified the people you need to define those rules. The rules need to be documented and readily available. I wouldn't say though that they need to be communicated to all stakeholders because those stakeholders should have been involved in creating them. Master data typically spans multiple departments in a company and all impacted departments need a voice to get the data right.

React to what exists, build in preventative measures to create quality data, and proactively maintain by monitoring and auditing to ensure all preventative measures are being followed

If you're not consolidating your MD, you don't really need to be reactive, i.e. no need for real curating processes. In most cases one can come up with a proactive DQ strategy, thus nearly eliminating the need for this step.

Use a master data reconciliation approach to merge or update conflicting entries, ensuring a single source of truth. Regular audits can help verify data accuracy, and setting up dispute resolution protocols prevents prolonged inconsistencies.

In my experience, creating a data culture is inherently transformative in an organization's MDM journey. It's a collaborative and shared mindset where ownership and accountability are every team member's responsibility. An organization that has a healthy data quality culture considers quality in data as a commitment and not a task.