Most content engines fail at the same point: month two. The team runs out of hooks. Not ideas, hooks. The opening sentence that earns the next 60 seconds of attention. Without a hook bank, every post starts from zero.
Ignis runs a hook bank per client. Roughly 200 reusable hook structures, organised by buying intent and tagged by topic. Built once in week 0, refreshed monthly with the winners and losers from the past 30 days of organic data.
The structure of the bank.
Layer one. The hook templates. About 30 reusable patterns. "Most [category] does X. The ones that win do Y." "Three things stop [outcome] from happening. Here's the order to fix them." "I spent $X on Y so you don't have to. Here's what worked." Each template is a fill-in-the-blank that produces a different post depending on which topic + audience pair you slot in.
Layer two. The topic taxonomy. The 10 to 15 topic clusters the brand owns. For VBC: trade marketing, product launches, supplier negotiation, retail floor optimisation, etc. For a B2B SaaS: customer acquisition, retention, expansion, pricing, hiring. Each topic is tagged by which buying intent it serves (awareness / consideration / decision / retention).
Layer three. The hook bank itself. Each row is template × topic × intent. "Most [trade marketers] do X. The ones that win do Y" becomes a slot-filled hook every time the team picks a new topic from the taxonomy. 30 templates × 12 topics × 4 intents = 1,440 unique hooks the team can pull from. We trim to the 200 that consistently perform.
Layer four. The performance feedback. Every shipped post is logged with its hook template + topic + save rate. After 30 days the bank is sorted by save rate. The top 20 percent get amplified into more variations. The bottom 20 percent get retired. The middle stays in rotation.
Why this matters. With a hook bank, daily output stays sharp through month six and year two. Without one, the team starts repeating themselves by month three, the audience notices, save rate drops, and the engine stalls. The hook bank is the difference between an engine that compounds and one that flames out.
Build the bank in week 0. Refresh monthly. Treat it as the operating asset, not a content library.