The first section of your DMP should consider how you are going to create the data, or if you are going to reuse an already existing dataset.
Type of Data
What will your research data be composed of? For example, will it be one-to-one interviews? focus groups? surveys? Or perhaps numerical/tabular data from a laboratory machine. Data could be received from a third party, e.g. clinical patient data from a hospital. Describe the nature of the data.
Volume of Data
Simply, what size do you envisage the dataset(s) to be. Of course, this can be tricky to estimate at the outset, and can always be adjusted as the data creation process progresses. That said, it is important to spend a little time here to get an approximation of the data set size, as the size of the data set can impact on a lot of the decisions down the line. Storing 1TB on data is relatively easy, storing 100TB is not. Preserving (long term) 1TB can be challenging, preserving 100TB can be impossible (or at least impractical).
List all file formats used and explain the rationale for choosing them. You may have chosen open data formats to increase reuse or remove the need for specialist (and potentially expensive) software to access them. Or you may have chosen the standard, discipline recognised formats for your files. Whatever the decision, list all formats and the rationale for choosing them. Here are some resources to help choosing your file formats...
If you are reusing existing data, consider the reuse licence terms. What are you legally able to use the data for, and can you make the data freely available after the research project. If you are licencing data for a specific timeframe or reusing freely available public data, much of these decisions are taken out of your hand (i.e. the file formats etc are already determined).