Data availability measures how available your data is when needed, either by your organization, partners, or by third parties (like government agencies for things like the GDPR).
Ideally, your data should be fully available and accessible to anyone that needs it at any time. However, roadblocks and obstacles can arise when accessing data, whether it be low data quality, poor labeling, or bad governance. Organizations use data management to work around these obstacles and make their data more available.
If your organization has several protocols and processes to access relatively low-level (not sensitive or restrictive) data, then it would not have optimal data availability. If it takes a very long time for data requests to be fulfilled, that is also a sign of poor data availability.
Data availability is one of the main objectives of any data management project. Your DM solution should address availability issues by providing metadata management, adding reference data and better labeling to aid in the search, profiling datasets to see which need to undergo preparation, and much more.
Redundancies and backups also help with data availability. If there is a failure in the primary system, people can work off the backup system until it is back online. This way, the data is always available regardless of system failures.
Needless to say, having your data be accessible to everyone that needs it is essential and a great advantage to any data-driven organization.