Once you have created some assumptions from what you know about your audience, and what the statistics tell you, do a test. Choose two or three times to send out the same message. Then check the numbers. Which time got the most opens, shares, and action. If there isn't much difference you may need to run a test on your headlines and message too.