fortnite-growth-lessons
Shared cross-project knowledge base of what actually worked or flopped in marketing/growth work (thumbnails, titles/tags, trailers, launch plans, patch-note tone, retention proposals) across every pro
SKILL.mduser-level-skills/fortnite-growth-lessons/SKILL.md
GitHubFortnite growth/marketing lessons learned
Same idea as ~/.claude/skills/uefn-lessons/ but for the growth/marketing side instead of Verse
code: this file is shared across every project set up with this kit, persists across sessions and
islands, and only earns its keep if growth-manager and the 8 fortnite-* skills are disciplined
about reading it before producing a new asset and adding to it once a real outcome is known.
Why this exists, and how it's different from just "doing the process again"
Each fortnite-* skill (thumbnail-pro, competitor-analyzer, etc.) already has a solid process
for producing an asset — but a process alone doesn't accumulate taste. Two thumbnail concepts can
both follow the "high contrast, dominant subject, 2-4 words" rules perfectly and still differ
wildly in real CTR. This file is where that gap gets closed: not "how to make a thumbnail" (the
skill's own job) but "here's a concrete case where this approach actually under/over-performed,
and why, as far as we could tell."
When to add an entry here vs. not at all
- Add an entry when a real outcome is known and attributable: the owner reports back CTR/ retention/Discover numbers after a change, a specific thumbnail/title/trailer approach got a clearly better or worse reaction than expected, a patch-note tone landed well or fell flat with the community, a launch tactic drove (or didn't drive) real traffic. Always note the actual signal it's based on (a number, a direct quote/reaction, a before/after comparison) — never file a lesson based only on "this felt like it worked."
- Don't add an entry for a concept/asset that was merely produced but never actually shipped or measured — that's not a lesson yet, it's just output. Wait for the real signal.
- If in doubt whether something belongs here vs. per-project memory (
growth-manager's ownmemory: project): here if it would change how you'd approach a DIFFERENT project's asset too; per-project memory if it's specific to this one map's audience/genre/history only.
How to add an entry
One line per lesson: what was tried, what the actual measured/observed outcome was, the likely reason if there's a plausible one — not speculative theory dressed up as a finding. Add under the closest matching category below (add a new one if none fits). Never delete or overwrite another entry without being sure it's wrong; if a later project seems to contradict an existing one, add a note next to it instead of silently replacing it — genre/audience-dependent contradictions are often the most useful entries here (e.g. "true for Horror maps, the opposite held for Tycoon").
Remove the "(example — replace)" placeholders once real entries exist; they only show the format.
Thumbnails & CTR
- (example — replace) A thumbnail with a single dominant character face outperformed a
multi-element "action scene" version on the same map by a wide CTR margin — the squint-test
rule in
fortnite-thumbnail-proheld up in this real case, not just in theory.
Titles, descriptions & Discover tags
- (example — replace) ...
Trailers & social hooks
- (example — replace) ...
Launch tactics
- (example — replace) ...
Patch notes & community tone
- (example — replace) ...
Retention/design proposals that were actually adopted and measured
Cross-references ~/.claude/skills/discover-retention/ for validated proposal outcomes — that
skill already tracks this category specifically; don't duplicate entries here, link to it instead
if a growth decision and a retention proposal are tied to the same real outcome.