The Influencer Marketing Stack: Four Skills, One Fingerprint

Influencer marketing is usually described as a relationships business. The four skills that cover it in our directory describe something else: a measurement pipeline with people at the ends.

Four stages, one author

The influencer family in our directory runs in strict order. audience-mapper — 9.0 — builds the audience profile first: who you are trying to reach, on which platforms, with what behaviors. influencer-discovery — 9.0 — takes that profile and produces a creator list, with vetting criteria rather than vibes. trend-spotter — 9.0 — surfaces the trends a brand could plausibly join, instead of the ones it is already late to. budget-optimizer — 9.0 — allocates spend across creator tiers, which is the stage where most campaigns actually fail.

What stands out in the evaluation data is not any individual score but the fingerprint: three of the four skills post identical dimension lines — trigger 9.5, structure 8.7, workflow 9.6, content 9.0, engineering 6.0, security 9.5. Influencer-discovery is off by a few tenths (trigger 9.0, structure 9.3, engineering 7.0). This is what a single competent author looks like in the numbers: one philosophy about how to structure a skill, applied repeatedly. Consistency at this level is rare — most families we review vary by a point or more across members, because they were written by different people at different levels of care.

The pipeline logic

The order is the argument. You map the audience before you search for creators, because a creator list without an audience definition is just a popularity contest. You check trends before you allocate budget, because a trend that has already peaked should change what a tier is worth. And you optimize budget last, because allocation is meaningless until the inputs above it are settled.

Each skill's trigger description follows the same discipline: "analyze my target audience," "find influencers," "what trends should my brand jump on," "allocate my influencer budget." The phrasings do not overlap, so the pipeline routes cleanly — a user mid-campaign gets the right stage without disambiguation. Workflow at 9.6 across the family reflects this: these are procedures with inputs, outputs, and hand-offs, not topic pages.

Downloads fall as strategy rises

The install counts run 12, 6, 5, 6 in pipeline order — the audience-mapper outpulls the trend-spotter more than two to one. That gradient shows up repeatedly in our marketing-category data: skills closer to the start of a workflow, where the work is concrete and the output is a list or a profile, get installed more than skills at the end, where the output is a judgment. People will let an agent build them an audience profile far more readily than they will let it decide where the money goes.

The shared weakness is the shared risk

The engineering dimension — 6.0 to 7.0 across all four — is the family's real limitation, and it lands exactly on the riskiest stage. Budget allocation is arithmetic over assumptions: creator rates, expected reach, tier mix, projected conversion. It is the one stage in this pipeline that should be a spreadsheet with named assumptions and sensitivity checks, not a well-written procedure. An agent that allocates budget from prose will produce confident numbers whose assumptions nobody can audit — and the failure will look like a bad campaign, not a bad skill.

Security at 9.5 across the family is solid; these skills touch no credentials and run no dangerous commands, which is what that score measures.

The stack is worth installing as a set — the fingerprint says the author designed them to compose, and the trigger descriptions keep the routing clean. But treat the budget-optimizer's output as a draft to interrogate, not a decision to execute. The skills encode the process of influencer marketing well. The judgment at the end is still yours, and the numbers' weakest dimension is the one that would make them checkable.