It's a delicate balance between refining the message to make it simpler & clearer, vs changing it because you're getting bored. At the beginning of the project it's probably important to try different approaches, but then stick with the message once you have it.
Many people would say that it's best to A/B test different messages in ads, and track which one works best. I'm finding it important to pitch the idea to 50 people first, get a feel for what resonates in the room and then test the messaging in an ad.