Overview
Importing data can be tricky because of the variances between systems so please take note of the following:
Use a Unique External Reference ID for each row
In most cases you will need to import several files to match up related information. For example you will first import the base action data (matter name, type, etc) and then import several related records for each action (parties, file notes, etc). If you use a unique external reference for the base data then you can pre-match this data for the corresponding file imports without having to use the ActionStep-generated IDs. This field may be called different things in different classes but it is usually fairly easy to identify (for example in the Action class it is called "External Reference #").
Save your file as a "Standard" CSV
Some programs offer different options for saving your CSV file e.g. "MS-DOS", "Mac", etc. Choose either a standard CSV format or, if not available, then "Windows CSV".
Make sure Dates are in YYYY-MM-DD format
This is the most reliable and most unambiguous format.
Break large CSV's into smaller files
Depending on server load at the time large files may cause the upload session to time-out so it is best to break large files into smaller chunks. Anywhere between 500 to 2,000 rows is usually okay but to be safe you might want to make them closer to 500 rows.
Remove CR/LF and other special characters from blocks of text
Sometimes these special characters can affect the formatting of the CSV file. For example a CR/LF in a text block could be interpreted as the end of the CSV row rather than a character within a data field. If you need to inspect the file use a hex editor to see the special characters.
Insert a Dummy Column at the end of the CSV file
CSV files vary on how they handle the delimiters on the last data field of each row. To prevent any potential problems add a dummy column to the CSV file and populate this column with some data e.g. "XXX" (don't leave it blank).