There are many reasons why research data should be managed 
First and foremost, researchers need to plan, design & organise their research process:
plan what data you need to collect, and how to collect it
decide what data will be used in your work and how will it be represented
plan how and where your research data will be stored for preservation and reuse
decide how your data may be made discoverable, accessed, shared and terms of re-use
These are the core elements of Data Management Plans (DMPs) and as such, are to be initiated at the start of any research project/undertaking. Well-developed DMPs increases research efficiency.
Creating a DMP is considered good research practice. Decisions made early on in the research project helps researchers save time, consider the necessary resources and costs. These will be required for funding/grant applications
Creating a DMP is considered good research practice. 
Decisions made early on in the research project helps researchers save time, consider the necessary resources and costs. These will be required for funding/grant applications
What is Data? 
Data is a scholarly product (like journal articles and books)
Research data are not a mere by-product of scientific research, nor a simple means to (article) publication. They often have a much longer shelf life than the scientific publications they underpin:
They constitute the evidence needed to verify and validate published claims.
They can be reused for follow-up or new research, for teaching, etc.
Therefore, research data should be cared for as first-class research objects. RDM is about exactly that. Ghent University
Many kinds of data created as part of a research project are subject to the same rights as literary or artistic work. Such items acquire rights like copyright or more general Intellectual Property rights when they are created. This gives the rights owner control over the exploitation of their work, such as the right to copy and adapt the work, the right to rent or lend it, the right to communicate it to the public and the right to license and distribute. UK Data Service
Funder and Publisher requirements
Meet funders and publishers research data requirements
“Planning how your data will be looked after – many funders now require data plans as part of applications” University of Oxford
Save time and resources
Research data management saves time and resources
Researchers will spend less time on data management and more time on research by investing the time and energy at the start of the project. University of Oxford
Data preservation and data security
Data preservation and data security as data, especially digital data, is fragile and easily lost. University of Oxford
Risk protection
Protect institutions from reputation, financial and legal risk. University of Oxford
Data Analysis
Good management helps to prevent errors and increases the quality of data analysis
Data Sharing
Research data management facilitates sharing of research data and, when shared, data can lead to valuable discoveries by others outside of the original research team
Research integrity
Well-managed and accessible data allows others to validate and replicate findings, and to ensure research integrity.
RDM is a due diligence practice that ensures the researcher retains transparency and accountability.