Glossary
Plain-English definitions for the terms behind LinkedIn and X automation: how feeds rank content, what actually triggers account restrictions, and what the AI-content vocabulary really means. Written by a team that builds these systems, not by a keyword tool.
Ranking
360Brew
LinkedIn's decoder-only foundation model for feed ranking, which judges a post's relevance from its text rather than from its engagement counts alone.
Safety
Account warmup
The gradual ramp-up of activity on a new or dormant social account to establish a normal baseline before running higher-volume automation.
Safety
Behavioral signature
The pattern of timing, IP source, device fingerprint, and action diversity that platforms use to tell automation from humans, often more telling than raw counts.
Reach
Dwell time
How long a viewer's screen stays on a post before scrolling away, used as a ranking signal because it is harder to game than likes or comments.
Safety
Rate limit
A platform cap on how many actions of a given type an account can take in a time window, often enforced as a rolling window rather than a daily reset.
Safety
Residential IP
An IP a consumer ISP assigns to a home connection. Platforms treat it as lower-risk than a datacenter IP because it matches ordinary human usage.
Metrics
Social Selling Index (SSI)
LinkedIn's 0-to-100 score rating an account on four areas: professional brand, finding people, engaging with insights, and building relationships.
AI Content
Voice matching
Reverse-engineering a person's writing style from samples so that generated drafts read as if that person wrote them, rather than as generic AI output.