With this few conversions, it's hard to make any judgement to what works and small tweaks in particular are basically impossible. Noise drowns everything. It's inviting to start getting fancy with A/B testing age ranges or gender targeting on FB ads, but unless I am mistaken about 230 conversions are needed to say A/B test male vs. female assuming it would give a 30% boost. If you know how to figure out the optimal targeting without spending ~$20000 on each test, let me know :-)
If I could get LTV up (reducing churn might be the lowest hanging fruit now), conversions up or write better ads, paid advertising could still be doable in the future.