Hey folks, it’s Saturday and I’ve been reflecting on how I structure outreach tests for new projects.
There are plenty of different angles people use - I’d say a dozen or more. The ones I often come back to are pain-driven, storytelling, and signal-driven. (If you have other favorites that work well, please drop them in the comments).
The way I run it: we take the role (or cluster of similar roles), randomize the sample, and launch all test sequences simultaneously on equal-sized cohorts. That’s the only way I’ve seen results stay clean.
Couple of things I’d love to hear your take on:
– From my experience, you need at least ~400 leads per angle to get a signal. 600+ is safer. How do you size your tests?
– When testing a new angle, do you stop at the first email + one follow-up (to deepen the hook), or do you also add a second follow-up with a fresh angle?
– Do you include a generic control sequence as a baseline? (Feels like without it, comparisons aren’t statistically valid).
Curious how others approach this. Would love to learn new tricks if you’ve seen something that works better.
