I've got a Facebook pixel that has about 100,000 conversions in it (been running for a year). We run all our campaigns on conversion optimizer. We have a pretty generic offer that doesn't require granular targeting.
I am considering launching a 1% lookalike audience on these 100,000 converters to see how it does.
My concern or suspicion is that while our pixel has 100,000 conversions, the pixel is likely already doing a good job at finding these "1% lookalike" people and serving ads to them. Creating my own lookalike and running a campaign to them is like manually doing work Facebook is already algorithmically performing for me. Does that make sense? Do most of you once you find something working, just let the pixel take over?
I would say to split test.
Your pixel has a solid amount of data for FB to optimise, but with stricter targeting, you may be able to squeeze more profits out of your campaign.
Not too familiar with this though, so maybe a veteran can chime in here.