Based on the volumes of the data and the coverage of data for your test cases you need to decide, on which tool the comparison should be done.
In any case, it is better to automate the process. But, if it is a onetime comparison and if the data is not more than 10-30 records, you can do the tests and verify them with the naked eye.
So, there are usually two options which are considered one is Excel and the other is the Database. Please note there are some automated tools and packages with the tools which do this, but assuming we don’t have any right now then we have to go with either Excel or Database.
When it comes to Excel, it is pretty easy, straight forward if you know the basic usage and the formulas and connecting to the data sources. But, we have a limitation of the volumes or the number of records to compare and if the volumes are high (which is common, now a days) then the excel starts to crash and you would have to redo the work over and over again.
So, the option left for us here is database. Now, it is not mandatory that your source or target or reference data is always a relational type. So, you need to load the various other data source formats into the database first.
Once you have all the related data to compare, you should start with the queries and applying the related transformation and filter conditions based on the Business Requirements.