See what the AI will write to your prospects, before it writes to them

Start with the first message on three or four different prospects: a profile very close to your audience, a borderline profile and one who came from a reaction to a post. If the messages look too similar, your instruction is probably too directive; if they go in every direction, your offer or audience lacks precision.

On Pro and Agency, then test two or three objections you often hear. You will see whether the AI answers the question before trying to move forward, and when it offers a slot. If a reply does not suit you, add an instruction, regenerate, then save.

Test follow-ups too: generate the second, then the fourth, to see how the sequence evolves and whether your instructions hold on later steps. On Agency, you can grant the test right to a client or a collaborator: they test with the settings of their own account, and nothing is sent.

Keep in mind that real prospecting adds a review before every send and takes the whole past conversation into account. The test shows you the direction, then the machine applies the same rules to every prospect.

A prospect who liked your post is not approached like a stranger found through search. The first message changes with the origin of the relationship: it can mention a reaction, thank someone for an invitation they sent, or go straight to the point with a first-degree connection. Testing each starting point shows you how your campaign will speak to each of your sources.

Six starting points are available. For a test even closer to reality, search for a prospect in your campaign and load them: their profile and context fill the simulation. Switching campaign clears that prospect, so a test never mixes two campaigns.