I've written three of my own and reviewed maybe fifteen others at this point. here's what separates the ones that land from the ones that get deferred.
the core mistake: people write a promo packet like a performance review. they list projects and responsibilities. that's not what it's for.
a promo packet is an argument. it needs to answer one question: why is this person operating at the next level right now, not 'will be ready soon.'
the structure that works
open with a 2-3 sentence thesis. literally: 'I am operating at senior level because X, Y, Z.' don't warm up to it. your manager's manager will skim this doc in 4 minutes.
then go impact-first, not project-first. instead of 'led the data migration project' write 'the data migration unblocked two product teams for Q3 and is estimated to save $40k/year in infra. I led it end-to-end including stakeholder alignment with finance.'
every bullet needs a signal, not just a task. signals are things like: made a decision others avoided, taught something that spread to the org, resolved ambiguity without escalating.
the peer feedback section matters more than you think
don't ask your friends. ask people who've worked on the hard things with you and who will be credible to your skip-level. three good quotes from the right people outweigh ten generic 'great to work with' lines.
what kills packets
claims without evidence. 'I have strong cross-functional influence' with no example is worthless. 'finance partnered directly with me to scope the budget model before involving my manager' is evidence.
also: scope creep in the writing. if the packet has 8 major projects in it your manager will wonder why you're not already at the next level. pick 3-4 that cleanly show the bar, and reference the others briefly.
one more thing: your manager submitting a promo packet is them going on record that you're ready. make it easy for them to defend it. the doc isn't for you, it's for the calibration room.