Reference
AI Social Media Marketing Glossary
Plain-English definitions for every term Content Drifter uses across the marketing surface, from copywriting frameworks (PAS, AIDA, BAB, FAB, PASTOR) to product features (Gold Slot, Voice-Match, branded carousel, closed-loop AI). Linked to the relevant feature page where one exists.
By Tayyab, Founder
Terms and definitions
- Gold Slot
A peak-engagement posting window drawn from published engagement research and your own results.
A Gold Slot is a specific time window on a specific platform where posts historically earn meaningfully higher engagement than the platform average. The free Gold Slot model uses industry-standard windows (e.g., LinkedIn 8 to 10am Tue to Thu local time); the personalized Gold Slot model on Pro+ re-ranks those windows using your own engagement history.
- Voice-Match Score
A 0 to 100 confidence score that an AI-generated post sounds like the brand.
Voice-Match Score compares each generated post against the brand-voice fingerprint captured in the Brand Kit (tone, vocabulary, hook patterns, banned phrases). Scores below 80 are flagged as drift. Most brands stabilize at 88 to 94 once the Brand Kit is dialed in.
- Closed-loop AI
AI that retrains generation on real-world engagement of its own past output.
A closed-loop AI content engine pulls per-post engagement (likes, comments, shares, reach) on a regular cadence and feeds the top + bottom performers back into the next batch's generation context. The result: month 2 outperforms month 1 because the AI is writing for an audience it actually knows. Generate-and-forget AI tools cannot do this.
- Brand Kit auto-fill
Scrape brand colors, logo, fonts, and tone-of-voice signals from a website URL in 30 seconds.
Brand Kit auto-fill takes any website URL and extracts the brand identity Drifter needs: primary colors, logo, font stack, target audience signals, and the tone of voice expressed in the existing copy. Cuts onboarding from 5 minutes to under a minute and prevents the typical "I never finished setting up the brand kit" abandonment.
- Branded carousel
A 5-slide Instagram or LinkedIn carousel with the brand colors, fonts, and logo applied per slide.
Branded carousels are multi-slide deck posts where every slide inherits the brand identity automatically, with no per-slide design work in Canva. Starter and above generate ready-to-publish carousels from a single prompt, with the Brand Kit driving visual consistency.
- PAS (Problem · Agitate · Solve)
A conversion-focused copywriting framework: name the problem, sharpen the pain, present the solution.
PAS is the most-used direct-response copywriting framework. It opens by naming a specific problem the reader has, agitates the cost of leaving it unsolved, then presents the solution as the obvious next step. Best for conversion-focused posts where the reader needs to be moved to action quickly.
- AIDA (Attention · Interest · Desire · Action)
A four-beat awareness framework: hook attention, build interest, create desire, demand action.
AIDA is the canonical advertising framework. The opening line earns attention; the next 2 to 3 sentences build interest by deepening context; the body creates desire by showing the transformation; the close demands a specific action. Best for awareness posts where the reader does not yet know they have a problem.
- BAB (Before · After · Bridge)
A transformation framework: paint the current state, the desired state, and the bridge between them.
BAB structures a post as a transformation arc. Before paints the reader's current frustrating state, After paints the desired state, Bridge names the specific lever that gets them from one to the other (the product, the insight, the playbook). Best for product-launch and case-study posts.
- FAB (Features · Advantages · Benefits)
A product-marketing framework that translates technical features into reader outcomes.
FAB chains each technical feature ("the product has X") to its user-facing advantage ("which means you can Y") to its emotional benefit ("so you feel Z"). Best for product feature posts where the reader cares about outcomes more than mechanics.
- PASTOR (Problem · Amplify · Story · Transform · Offer · Response)
A long-form persuasion framework with built-in narrative and urgency.
PASTOR is the long-form sibling of PAS, with a story beat for credibility, a transformation beat for vivid outcome, and an explicit offer + response (CTA) at the end. Best for launch-week urgency posts and limited-time offers.
- STAR (Situation · Task · Action · Result)
An industry copywriting framework that turns customer wins into a clean narrative arc.
STAR is a widely-used storytelling framework that structures a customer story by Situation (the context), Task (what they needed to accomplish), Action (what they did), and Result (the measurable outcome). It is a useful pattern for case-study and testimonial writing. Content Drifter's engine ships five built-in frameworks (PAS, AIDA, BAB, FAB, PASTOR); STAR is defined here as a general industry concept rather than a Drifter framework option.
- 4Ps (Promise · Picture · Proof · Push)
An industry positioning framework that anchors a brand promise with a vivid picture, proof, and push to action.
4Ps is a general copywriting framework that opens with the brand promise (the headline transformation), paints a picture of life with that promise fulfilled, proves it with evidence (data, customer quote, demo), and pushes the reader to the next step. It is a useful pattern for top-of-funnel positioning. Content Drifter's engine ships five built-in frameworks (PAS, AIDA, BAB, FAB, PASTOR); 4Ps is defined here as a general industry concept rather than a Drifter framework option.
- Algorithm-aware engine
A content generator that adapts copy and cadence to each platform's current ranking signals.
An algorithm-aware engine maintains a per-platform profile of what the current ranking algorithm rewards (hook patterns, length, hashtag density, link placement, posting frequency) and feeds those signals into the generation prompt. The profile is refreshed on a regular cadence: a daily algorithm canary watches aggregate engagement and updates the profile when the wind changes.
- Auto-Pilot
Hands-free scheduling and publishing that respects blackout windows and re-optimizes the next 7 days overnight.
Auto-Pilot is the fully-automated mode of the scheduler. It schedules every approved post at the highest-leverage Gold Slot, avoids any blackout windows the brand has set (e.g., weekends, holidays), and re-optimizes the next 7 days nightly using the latest engagement data. Pro+ feature.
- Blackout window
A time range when the scheduler will never publish, even if a Gold Slot falls inside it.
Blackout windows let brands enforce posting boundaries. Common patterns are "weekends off," "no posts during the daily 9am all-hands," "no posts during the 2-week holiday shutdown." Auto-Pilot and every scheduler tier respect blackout windows automatically.
- Marketing campaign engine
A tool that generates a multi-week campaign blueprint with intent-tagged content slots and per-slot copywriting frameworks.
The marketing campaign engine takes a 60-second brief (goal, horizon, target audience, primary platforms) and outputs a 7- or 30-day blueprint with platform mix, posting cadence, and a content slot per post. Each slot is tagged with a narrative intent (awareness, conversion, urgency, transformation, etc.) and the engine picks the matching copywriting framework (PAS / AIDA / BAB / FAB / PASTOR) automatically.
- Repurpose engine
A tool that turns one long-form piece (blog or YouTube link) into 10 to 20 platform-native short-form posts.
The repurpose engine accepts any URL (a blog post or a YouTube link) and atomizes it into platform-native posts for every connected network in 30 seconds. The pillar piece earns the search rank; the atomized posts earn the social distribution.
- Hook library
A curated collection of 96 battle-tested viral opening-line patterns, browsable by category and platform.
The hook library categorizes opening lines into question hooks, statistic hooks, story hooks, contrarian hooks, list hooks, and 6 more, each with platform-specific examples. The AI hook rewriter (Pro+) can apply any pattern to an existing draft in one click.
- Trends feed
A live feed of newsjacking opportunities sourced from Reddit, RSS, YouTube, and Mastodon.
The trends feed surfaces emerging conversations in the user's niche, refreshed every 30 minutes from public sources. Each item ships with a "Generate post on this" handoff that drops the topic + context into the content generator with the brand voice already applied.
- Content preset
A saved bundle of brand voice, target audience, hashtag rules, forbidden words, and sample posts that gets applied to every generation.
Content presets are the reusable layer of brand voice. Once configured, every batch generation can apply the preset with one click, with no re-typing the brand voice into the prompt. Pro+ users can maintain multiple presets (e.g., one per content pillar or per persona).
- Approval workflow
A staging area where generated posts await human approval before publishing.
The approval workflow holds every generated post behind a per-post gate until a human approves it: the scheduler skips unapproved posts and the publish API refuses them. Per-platform validation runs at review time (character limits, image dimensions, link rules) so nothing publishes that a platform would reject, and you stay in control of everything that goes out under your brand.
- Per-platform control
The ability to specify exactly how many posts to generate per platform per batch.
Per-platform control gives the user precise authority over batch composition: "2 LinkedIn posts, 3 X posts, 1 Instagram carousel." Replaces the default "generate one of everything" behavior that wastes the AI quota on platforms the brand does not actively use.
- Multi-platform publishing
Auto-publishing the same approved post (with platform-native variants) to multiple networks from one workflow.
Multi-platform publishing connects every supported network via OAuth (Instagram, TikTok, X, LinkedIn, Facebook, Threads, YouTube, Pinterest and Bluesky) and publishes natively, with no copy-paste between tabs. Every network is available on every plan. Each platform receives its tailored variant (length, hashtags, image dimensions) automatically.
- Auto-publishing
Native, OAuth-based publishing that pushes approved posts live at their scheduled time, hands-free.
Auto-publishing eliminates the manual "open the platform, copy the post, click publish" loop. At the scheduled time, the publisher posts directly via the platform's native API. Posts are tagged with a unique identifier so the analytics worker can track them through the wild.
- Per-post analytics
Real engagement metrics (likes, comments, shares, reach) pulled per post on a regular cadence.
Per-post analytics are pulled every 30 minutes (more frequently in the first 6 hours after a post when the engagement curve is steepest). The data feeds two surfaces: a What's Working dashboard for human review, and the closed-loop AI for next-batch retraining.
- On-demand credits
Pay-as-you-go AI generation credits that supplement the monthly plan quota.
On-demand credits never expire and kick in automatically when the monthly plan quota runs out. Sold in 50 / 200 / 500 packs. Useful for one-off campaign bursts (a launch week) without committing to a permanent plan upgrade.
- Content calendar
A visual planner that shows every scheduled, draft, and published post in month and list views.
The content calendar is the single-pane view of every post across every platform. Drag-and-drop rescheduling, filter by platform / status / brand, search by content. Replaces the "guess what's posting tomorrow" anxiety that kills consistency.
- Smart scheduling
AI-powered scheduling that recommends the single best moment to post based on the content + audience.
AI smart scheduling reads the content (topic, intent, length), the audience (target persona, time zone), and the platform (current ranking signals) and recommends one optimal posting time per post, not a generic peak window. Sits between Gold Slot (industry-standard windows) and Auto-Pilot (fully automated cadence).
- Brand voice
The consistent tone, vocabulary, and stylistic patterns that make a brand's content recognizable.
Brand voice is the structured artifact that lets multiple writers (or one AI) produce content that consistently sounds like the brand. A working brand voice has explicit rules: tone (e.g., direct vs friendly), vocabulary (preferred terms + banned phrases), hook patterns (the kinds of openers the brand uses), and CTA style.
- Target audience
The specific reader the brand is writing for, defined narrowly enough to make every content choice obvious.
Target audience is the named, narrow definition of the reader. "Series A SaaS founders in the US, $1M to $10M ARR, 5 to 25 employees" is useful; "B2B leaders" is not. The narrower the definition, the more obvious every copy / framework / channel choice becomes.
- Trial (in-product)
A time-bounded free trial of a paid plan, optionally with a card on file.
In Drifter, "trial" refers to the time-bounded free Pro trial new customers can start at /onboarding/trial. The trial requires a card on file but does not charge during the trial window. Paid plans separately: the Starter plan has a 7-day money-back refund window (see /terms). Pro and Agency rely on cancel-anytime billing (no money-back window).
- Aha moment
The specific in-product moment when a new user perceives the product's value for the first time.
In Drifter, the canonical aha moment is the first batch of generated posts in the user's brand voice, typically achieved within 60 seconds of completing the brand-kit auto-fill. The onboarding flow is engineered to compress time-to-aha as aggressively as possible.
- OAuth scope
A specific permission a platform grants a tool, for example "read posts," "publish posts," "read analytics."
OAuth scopes are the granular permissions Drifter requests when a user connects a platform. We request the minimum scopes needed for each capability (publish + read analytics). Full scope list per platform is documented at /security.
- Engagement rate
Total engagement (likes + comments + shares) divided by reach (or follower count) for a post.
Engagement rate is the canonical post-quality metric. Two flavors: ER-by-reach (engagements / reach, more honest) and ER-by-followers (engagements / follower count, more commonly cited but less honest). 2 to 5% ER-by-followers is typical for B2B; 5 to 10% for niche B2C; >10% is exceptional.
- Algorithm canary
A daily process that watches aggregate engagement across the user base for early signs of platform algorithm shifts.
The algorithm canary aggregates engagement signals across all Drifter users on each platform daily, looking for statistically significant moves in hook performance, length preference, hashtag effectiveness, etc. When a shift is detected, the per-platform generation profile is updated automatically, so the user's next batch already adapts before the news cycle reports the change.
- White-label (agency)
Using Drifter as the agency's own back-end production layer, with the content delivered as the agency's work.
White-label is the general agency model of using a tool as the back end while the finished work is presented as the agency's own. With Drifter (Agency plan), the agency runs the engine and hands the resulting posts and carousels to clients as their deliverables. Fuller white-label support (a per-client login on the agency's own domain, plus white-labeled client reports) is on the roadmap, not yet built.
- Image Studio
The unified creator for branded graphics: pick a layout family, edit the payload, re-render for free.
Image Studio replaced the three separate image tabs with one creator. You choose from 32 layout families (quote cards, before/after, product hero, photo spotlight and more), and the generated image stays editable: every render stores the family, payload, preset and brand kit that produced it, so changing a headline re-renders deterministically without spending a new image credit.
- Media library
The workspace store for every uploaded and generated asset.
The media library holds uploads and generated images in one place, so an asset can be reused across posts and platforms instead of being re-uploaded per post. Deleting an item removes the underlying blob as well, which keeps storage aligned with what the account can actually see.
- MCP server (AI agent access)
A Model Context Protocol endpoint that lets an AI assistant generate, schedule and publish on your behalf.
MCP is an open standard for giving AI assistants tools. Content Drifter exposes its own MCP server, so an assistant such as Claude can create posts, put them on the calendar and publish them using your account, subject to the same plan limits and approval rules as the dashboard. Access is authenticated with an API key, not a shared password.
- API key
A revocable credential that authenticates programmatic access to your workspace.
API keys authenticate the MCP server and any direct API use. Each key belongs to one workspace, can be revoked without affecting the others, and carries the plan limits of the account that issued it. Keys are shown once at creation and stored hashed, so a lost key is replaced rather than recovered.
- Team seat
A member of a shared workspace who can create, edit and schedule against the owner's plan.
Seats make a workspace shared rather than personal. Members work against the owner's quota and brand kits instead of holding their own subscription, which is what lets an agency or a small team run one calendar together. Seat counts come from the plan (Pro and Agency include more than one).
- Document post (LinkedIn carousel)
LinkedIn's swipeable multi-page format. It takes a PDF, not images.
A LinkedIn carousel is a document post: you upload a single PDF and LinkedIn renders each page as one swipeable slide. This is why an image upload produces a multi-image post rather than a carousel, and why carousels are designed at 1080x1350 (4:5) and exported to PDF. Instagram carousels take images instead, at the same dimensions.
- Content queue
A running order you add posts to without choosing a time; the next open peak-engagement slot is chosen for you.
A content queue removes the smallest and most common friction in publishing: deciding when. You add a post, and it takes the next open Gold Slot, skipping blackout windows and any slot already holding a post. The number worth watching is not how many posts are waiting but the DATE the queue runs dry, because a date is a deadline and a count is trivia. Content Drifter also draws the gaps in the queue as rows of their own, since a chronological list cannot show silence: the space between two posts has nothing to render.
- First comment (link placement)
Publishing a post without its outbound link and posting the link as the first comment, so the post keeps the reach a body link costs it.
Several networks measurably deprioritise posts carrying an outbound link: on LinkedIn that costs about half the reach, and X applies a substantial penalty of its own. The workaround is to publish the post without the link and add the link as the first comment. Most schedulers attach a first comment but leave the link in the body, which keeps the full penalty; the body has to be rewritten for the technique to do anything at all.
- Content categories (rotation)
Buckets with weights that decide the proportion of your feed each kind of post occupies.
Content categories turn "I should post more educational content" into something the scheduler enforces. Each category carries a weight, and the queue draws from them in that ratio. The part that separates a good implementation from a naive one is interleaving: a weighted picker can honour a 3:1 ratio while publishing four promotional posts in a row, which is the lumpy feed the mix existed to prevent. Content Drifter also measures the gap between the mix you set and what you actually published, and says so plainly.
- Evergreen recycling
Republishing a post that still holds up, into a different part of the week from its last airing.
Recycling keeps a feed alive during the weeks somebody does not show up, which are the weeks they are most likely to churn. The detail that matters is timing: republishing into the next available slot regardless of when the post last ran shows it to roughly the same slice of the audience twice, which is the one outcome recycling should avoid. Posts also retire after a finite number of airings, because without a cap an evergreen library quietly becomes a loop and the owner is the last to notice.
- Drip campaign
One source URL released as many posts spread over weeks rather than published all at once.
A drip takes content that already exists (a blog post repurposed into 10 to 20 platform-native posts) and spreads it across peak-engagement slots over weeks, at most a couple a day, with networks interleaved so consecutive posts prefer different platforms. Because the posts are already generated, dripping them costs no additional AI generation: it is placement of output already paid for.
- Click attribution (per post)
Counting link clicks against the individual post that earned them, rather than against the URL.
Most schedulers count clicks per link, so the same URL reused across five posts collapses into one number nobody can attribute. Content Drifter mints a separate tracked short link for every post and network combination, so a click resolves to one specific post on one specific network, and analytics can rank by people who actually arrived rather than by likes. Bot clicks are counted separately and excluded, because link previewers fetch every shared URL.
- Hashtag performance
Measuring which of your hashtags actually help, against your own posts that do not carry them.
A hashtag reported as "3.2% engagement" says nothing without knowing what is normal for that account, so the useful figure is a contrast: do posts carrying this tag beat the ones that do not. The harder discipline is refusing to rank a tag at all until it has appeared on enough measured posts, because below that a single lucky post decides the number and the tool has laundered luck into advice.
- Bulk CSV import
Moving an existing content queue in from another scheduler in a single upload.
The biggest practical blocker to switching schedulers is that the queue lives somewhere else. A useful importer recognises the column names another tool exports, validates every row against the same rules the publisher enforces (so an import cannot create a post that fails weeks later), accepts the good rows while reporting the bad ones by line number, and schedules rows that carry no date rather than rejecting them.
Frequently asked questions
What does engagement rate actually measure?
Total engagement (likes, comments and shares) divided by reach, or by follower count where the network does not expose reach. Which denominator you use moves the number by an order of magnitude, so a rate is only comparable against another rate that used the same one. The full entry is in the list on this page.
Where do these definitions come from?
From the specs and behaviours our own product implements rather than from a keyword list. 35 of the 50 terms link straight to the feature or page where the term does real work, so you can check a definition against the thing it describes instead of taking our word for it.
Do you ever remove a term from this glossary?
No. This glossary is append-only. Once a term is published, people and search engines have linked to it, so deleting one would break those links; a term we stop using has its entry rewritten to say so and to point at the replacement instead of turning into a 404.