Use equal segments or a small sample?
You can use equal segments or a sample, depending on your business need.
Before you segment a database, you must first have a segmenting field in the database. Because you cannot change the field type of an existing database field, you must create a new segmenting field.
You can divide the database into equal segments or you can sample the database. Use the Calculate button to see how many segments or segment sizes (number of contacts) will be made. The Calculate button calculates the number of segments or the number/percent of contacts per segment based on quantities that you specify.
Equal segments
Equal segments divide the database into two or more divisions of equal size. You might want to do this to divide a large email into smaller sections to be sent at different times. Dividing your emails can help regulate the flow of responses from your email to your web site.
Suppose you send a email to 100,000 contacts offering each a free gift for visiting your web site. If you send the email to all the contacts at once, your web site can be flooded with responses. If you segment the database into 10 equal segments and send the email to one segment per hour for ten hours, the flow of the responses is more manageable.
To divide the database into equal segments, you can divide by the number of segments that you want, or divide by the number of contacts you want in a segment.
Small sampling
Divide a database into small samplings to test responses to offers, subject lines, creative elements, and so on. You can also create samplings by sending to groups within your database.
For example, you might think that contacts are more likely to open an email with a subject line of 'Thanksgiving Sale' than that of 'Autumn Sale' or 'Fall Sale'. Using segmenting, you can send the email to different segments of the database to test the success of the different subject lines.
In this example, you can choose to send the three test emails, one with each of the above subject lines, to only small segments of a large database. You can then review the results of the emails and send the remaining contacts a email using the most successful subject line.