Glossary

AI search fundamentals

AEO

Short for Answer Engine Optimization — the practice of making your brand and content more likely to be named, recommended, and cited by AI answer engines like ChatGPT and Google AI Overviews. Where SEO optimizes for a rank on a list of links, AEO optimizes for presence inside a generated answer. See AEO, GEO & AI search explained.

GEO

Short for Generative Engine Optimization — the same goal as AEO (getting picked by the AI), described from the angle of the generative technology doing the answering rather than the user-facing outcome. The two terms are used interchangeably in practice; these docs mostly say “AI visibility.” See AEO, GEO & AI search explained.

AI search / answer engine

An answer engine is any system that responds to a question with a written answer instead of a list of links — ChatGPT, Google AI Overviews, Perplexity and others. “AI search” and “generative search” describe the same category. See AEO, GEO & AI search explained.

ChatGPT

OpenAI’s conversational AI assistant, and one of the two answer engines OptimizeCamp tracks today. When it’s configured with a live model key, OptimizeCamp asks it your tracked prompts directly and parses the real response. See Answer engines we track.

Google AI Overviews

The AI-generated summary Google shows at the top of some search results, above the classic blue links. The second engine OptimizeCamp tracks. Overviews don’t appear for every query — a missing one is a legitimate result, not an error. See Answer engines we track and Troubleshooting.

How AI answers are measured

Mention

A place your brand’s name (or one of its aliases) appears in the text of an AI answer. A mention doesn’t require a link — just being named counts. See Mentions, citations & share of voice.

Citation

A link an AI answer includes to back up what it says. Being cited means a specific URL of yours was used as a source — a stronger, more durable signal than a text mention alone. See Mentions, citations & share of voice.

Sentiment

Whether the text around a mention reads positive, neutral, or negative. Real answers get judged by a small language model, and each mention’s sentiment is stored with the snippet it came from, so it’s always auditable. See Mentions, citations & share of voice.

Position & prominence

How early your brand is named relative to every other brand in the same answer — first named is position 1. Prominence matters because being named first reads as a stronger endorsement than being named fourth. See Metrics & the Visibility Score.

Mention Rate

The percentage of scoped runs where your brand was mentioned at all — text-level presence. One of the four ingredients in the Visibility Score. See Metrics & the Visibility Score.

Citation Rate

The percentage of scoped runs where one of your pages was cited as a source — link-level presence. See Metrics & the Visibility Score.

Share of Voice

Your brand’s share of total mention volume across you and every tracked competitor, expressed as a percentage — the headline competitive number. A citation-based version (Citation Share of Voice) exists too. See Mentions, citations & share of voice.

Visibility Score

The single 0–100 number that summarizes your AI visibility — a weighted blend of Mention Rate, Citation Rate, Position Score, and Sentiment Index. Computed identically for your brand and every tracked competitor. See Metrics & the Visibility Score.

The measurement objects

Prompt (tracked prompt)

A specific question OptimizeCamp asks each AI platform on your behalf — e.g. “best project management tools for small teams” — and then checks the answer for your brand. See Managing tracked prompts.

Run (prompt run)

One execution of one prompt against one platform — the atomic unit of measurement. A run either produces an answer or doesn’t (Google AI Overviews sometimes doesn’t trigger, and that’s valid data too). See Prompts, runs & answers.

Batch (run batch)

A group of runs executed together — every active prompt × every tracked platform — kicked off by your onboarding baseline, a scheduled cadence, or a manual “Run tracking now.” Only one batch runs per project at a time. See Prompts, runs & answers.

Answer

The actual text (and any citations) an AI platform returned for a run, stored verbatim so you can read exactly what the AI said — not a summary or a score, the real response. See Reading an answer.

Brand profile

The record of who you are for measurement purposes: brand name, aliases, domain, description, category, audience, and topics. Everything downstream — matching, scoring, prompt generation — depends on getting this right. See Your brand profile.

Aliases

Alternate names or spellings that count as a match for your brand or a competitor when scanning an answer — e.g. “OptimizeCamp” and “Optimize Camp.” Matching is exact and case-insensitive, so a real mention written under a missing alias goes uncounted. See Your brand profile and Troubleshooting.

Competitor

Another brand tracked alongside yours and scored on identical rules for head-to-head comparison. Add one manually, or promote a brand the AI keeps naming that you haven’t tracked yet. See Competitors.

Prompt version (superseded)

Editing a tracked prompt’s wording doesn’t overwrite it — it creates a new version and retires the old text to superseded status, which keeps its history on the trend chart but drops it from active tracking. See Managing tracked prompts.

Trend break (annotation)

A dashed marker on a trend chart flagging that measurement conditions changed on that date — a prompt was edited, or your brand aliases changed — so a methodology shift isn’t misread as a real change in visibility. See The Visibility dashboard.

Fixing and creating content

Content inventory

The library of every page OptimizeCamp knows about for your project — pulled in from WordPress, imported by URL, or pasted directly. Nothing gets measured or fixed until it’s in this library. See Your content inventory.

QuickScore

A free, instant 0–100 heuristic computed the moment a page is imported or synced — structure, a clear definition up top, FAQ presence, structured data, length, and scannability. No model call, no credit spent. See Your content inventory.

Priority Score

The free, always-on 0–100 score that ranks which existing page to fix first, blending how much opportunity a page’s matched prompts still hold, whether it’s losing citations, how thin its structure is, and how stale it is. See The Optimization Queue.

Audit

The deeper, credited scoring pass you run inside the editor — parallel engines that check accuracy, citability, authority, AI coverage, and (when data exists) AI visibility, each producing concrete, click-to-accept fixes. See The audit & editor.

Engine

One of the audit’s independent scoring dimensions — the Accuracy engine, the Citability engine, and so on. Each runs on its own, contributes its own score to the composite, and can degrade gracefully without failing the whole audit. See The audit & editor.

Fix / issue

A specific, click-to-accept change the audit surfaces on a page — a rewritten sentence, a missing FAQ block, a citation to add — tagged critical, warning, or suggestion. Accepting one moves your score toward full recovery. See The audit & editor.

Gap (content gap)

A topic or question where competitors get mentioned or cited and you don’t even have a page — the raw material Generate turns into a brief. See Content gaps.

Opportunity Score

The 0–100 score that ranks gaps worth pursuing — how many losing prompts sit in the cluster, how often competitors get cited there, how high-intent the prompt category is, and how close the topic already is to something you’ve published. See Content gaps.

Brief

The editable plan for a new article, generated automatically from a gap’s evidence — title, outline, target queries, entities, “citations to beat,” sources, and voice notes — reviewed by you before a single word gets written. See Briefs.

Draft

The AI-generated article that comes out of an approved brief — grounded in retrieved sources, fact-checked, and landing pre-audited in the editor for human review before it ever reaches WordPress. See Writing articles.

Proving the loop worked

Verification state (verification window)

The window that opens the moment you publish a fix or a new page: the next tracking runs on that page’s matched prompts (or 14 days, whichever comes first) are what the eventual Impact Card compares against. See Impact Cards & wins.

Impact Card

The before/after receipt for one publish — did you gain a citation, how did mention rate move, how did visibility on that page’s matched prompts move — ending in a verdict of Citation gained, Observed change, No change, Pending, or Inconclusive. See Impact Cards & wins.

Action Center

Three lanes across the product: workflow ready to finish, evidence-backed opportunities, and measurement or results. Each card pairs a signal with a concrete next action without comparing unlike scores. See The Action Center.

Earned visibility & playbooks

Source (earned source)

A third-party domain AI cites on your prompts that’s neither yours nor a tracked competitor’s — a review site, community, directory or roundup. Rolled up one row per domain on the Sources screen.

Source type

How AI tends to use a source domain — one of community, review, directory, reference, or editorial (the default) — which decides the kind of playbook a source gets. See Sources.

Presence

Whether your brand actually appears on a cited third-party page — Present, Absent, or — (not checked yet). Drives the “You’re absent from” count and the type of play suggested. See Sources.

Play (earned-media playbook)

A typed, tracked off-site action against one source — review campaign, join the conversation, claim your profile, build a reference entry, pitch the roundup, or defend your listing — moving from suggested through verification to a won or no-change verdict. See Earned-media playbooks.

Earned Impact Card

The before/after receipt for a play, written once its verification window closes — did the domain start citing you, did your presence flip, did you gain prompt coverage. Correlational by design. See Earned-media playbooks.

Plans & usage

Audit credit

What your plan spends each time you run the deeper audit engines on a page inside the editor. Priority Score, QuickScore, and the Optimization Queue itself never cost a credit — only the full audit does. See Plans, quotas & credits.

Article credit

What your plan spends each time you generate a full draft from an approved brief. Section-level regeneration inside an existing draft is free; a full regenerate spends a new credit. See Plans, quotas & credits.

Cadence

How often your tracked prompts run automatically, set per platform by your plan — weekly on Starter; Google AI daily and ChatGPT 3×/week on Pro and Agency — plus a monthly allowance of on-demand “Run tracking now” batches for when you don’t want to wait. See Project settings and Plans, quotas & credits.

On-demand run

A manual “Run tracking now” batch, beyond your scheduled cadence — each spends one on-demand-run credit (Starter 2, Pro 10, Agency 40 per month). Scheduled and baseline runs are free. See Plans, quotas & credits.

AI outreach draft

AI-written outreach copy for a play, metered by a separate monthly allowance (Starter 5, Pro 20, Agency 50). Keyless template drafts are free. See Plans, quotas & credits.

History window

How far back trend charts let you browse — 6 months on Starter, 18 on Pro, unlimited on Agency. Older data is locked, never deleted; upgrading reveals it instantly. See Plans, quotas & credits.

Baseline (baseline run)

The very first run batch a project ever executes, during onboarding — every prompt you selected, asked of every platform. It’s the starting line every later run gets compared against. See Quickstart: onboarding.