> ## Documentation Index > Fetch the complete documentation index at: https://docs.snowdoughnut.com/llms.txt > Use this file to discover all available pages before exploring further. # Glossary > Definitions for terms used across this knowledge base. Linked from their first mention in other docs. Terms are listed alphabetically. Where a term has its own full page, this entry gives the short version and links out. ### Ad group Google's term for a group of ads that share the same customer profiles, keywords, and bids within a campaign. Equivalent to a Meta [ad set](#ad-set). See [ad set / ad group / variant naming conventions](/ad-set-group-naming-conventions). ### Ad set Meta's term for a group of ads that share the same customer profile, budget, schedule, and placements within a campaign. Equivalent to a Google [ad group](#ad-group). See [ad set / ad group / variant naming conventions](/ad-set-group-naming-conventions). ### AI Overview Google's model-generated answer shown above the results list, assembled from a small number of sources it cites. One of the surfaces [GEO](#geo) is concerned with. See [SEO and GEO concepts](/seo-and-geo-concepts). ### Article brief The fill-in workbook that defines what a single article has to answer, what new information it carries, who it is for, and how it is structured, before anyone writes it. Distinct from a [copy brief](#copy-brief), which governs ads, emails, and captions. See [article brief](/article-brief) and [article brief concepts](/article-brief-concepts). ### Asset register The list of every approved brand file with its formats, location, owner, approved uses, and review date. It exists so nobody has to ask whether a file exists or whether they're allowed to use it. See [brand assets](/brand-assets). ### Authority area A subject where a team can answer a question better than anyone else, on the basis of data others do not have, experience others have not accumulated, or a position others hedge on. Each authority area gets one [hero article](#hero-article). See the [article library overview](/article-library-overview). ### Boilerplate The 40 to 60 word third-person description of the company used on press releases, About pages, and directory listings. The longest item in the message set, and the one most often left to drift. See [messaging](/messaging). ### Campaign The top-level container for ads in a platform account: it sets the objective, budget type, and funnel position that everything underneath it inherits. See [campaign naming conventions](/campaign-naming-conventions). ### Canonical URL The address declared as the authoritative version of a page, so near-duplicate versions do not compete with each other. Usually the page itself. See [article standards](/article-standards). ### Chunk A passage of a page that a retrieval system pulls out and works with on its own, rather than the whole page. Sections are the practical unit, which is why each one is written to answer its own heading. See [SEO and GEO concepts](/seo-and-geo-concepts). ### Clear space The protected empty area around a logo that nothing may enter: no text, no other marks, no page edge. Defined as a proportion of the logo itself, usually a letterform's height, so one rule holds at every size. See [logo usage](/logo-usage). ### Co-branding Placing the brand's mark alongside a partner's. Relative size and order are read as status whether or not that was intended, which is why the rules fix both. See [logo usage](/logo-usage). ### Competitive frame The category a buyer mentally files a brand under, which decides what it gets compared against and on which attributes. Chosen rather than inherited. See [positioning](/positioning) and [brand strategy concepts](/brand-strategy-concepts). ### Content decay The gradual decline in traffic, rankings, and citations as an article's facts age, the question shifts, or someone publishes something more current. It does not announce itself, which is why articles carry a scheduled review date. See the [article maintenance process](/article-maintenance-process). ### Contribution margin What's left from revenue after every cost that scales with an order, including advertising: cost of goods, shipping, fees, discounts, and ad spend. A common [north star metric](#north-star-metric) for ecommerce businesses, since it's hard to improve by spending recklessly. See the [ecommerce metrics starter pack](/ecommerce-metrics-starter-pack). ### Copy brief The fill-in workbook that defines what a specific piece of writing needs to say and do, across whatever channel it runs on: email, SMS, paid social, search, or elsewhere. See [copy brief](/ad-copy-standards) and [copy brief concepts](/copy-brief-concepts). Use "copy briefs" only when referring to more than one. ### Counter-metric Watched alongside a [north star metric](#north-star-metric) to catch damage from optimizing too hard for it. Doesn't need its own target, only visibility in the same report. See [reporting framework concepts](/reporting-framework-concepts#the-north-star-metric). ### Creative brief The fill-in workbook that defines what a piece of creative needs to say and do before production starts: objective, customer profile, offer, and format. See [creative brief](/creative-brief) and the [creative brief process](/creative-brief-process). Use "creative briefs" only when referring to more than one, such as the [creative brief library](/creative-briefs). ### Creator The person who produces the creative work. This is the designer, art director, photographer, videographer, illustrator, copywriter, or other maker who takes the brief and turns it into the actual asset. ### Customer profile A defined segment of people an ad, campaign, or landing page is built for, described by shared traits, behavior, or list membership. Use "customer profile" for this concept everywhere, not "audience." See [customer profile brief](/customer-profile-brief), [customer profile naming conventions](/customer-profile-naming-conventions), and the [customer profile library](/customer-profile-library-overview). ### Data dictionary The fill-in document that defines a business's metrics, formulas, sources of truth, and report registry. See the [marketing reporting data dictionary](/marketing-reporting-data-dictionary). Not to be confused with a naming-convention [dictionary](#dictionary), which is a table of approved identifier values rather than metric definitions. ### Dictionary The table of approved values for a naming identifier (for example, every valid objective or offer code), built once per category and reused in every name assembled from it. Each [naming convention](#naming-convention) doc holds the dictionary for its level. ### E-E-A-T Experience, expertise, authoritativeness, and trust: the qualities Google's rater guidelines describe for assessing content quality. Not a score in the ranking system, but a useful shorthand for why a named author with first-hand experience carries more weight than an anonymous post. See [article brief](/article-brief). ### Elevator pitch The spoken 30-second version of the value proposition: customer, problem, solution, differentiation. One of the four lengths in the message set. See [messaging](/messaging). ### Entity A thing an engine recognizes as a distinct object with facts attached: a company, a person, a product, a place, a concept. Entity authority is built by describing the same entity consistently everywhere it appears, on your own site and on the profiles you control elsewhere. See [SEO and GEO concepts](/seo-and-geo-concepts). ### Experiment ID The short, structured identifier assigned to a test when its hypothesis is approved, appended to the names of everything that belongs to that test. See the [experiment ID SOP](/experiment-ids) and [experiment ID workbook](/experiment-id). ### Flat file The single spreadsheet holding every metric, for every channel and reporting period, stored as one row per period, channel, and metric rather than a wide table. Reports and dashboards are views generated from it. See the [flat file template](/flat-file-template). ### GEO Generative engine optimization: the work of getting a page retrieved, trusted, and cited when a model writes the answer rather than listing links. Overlaps heavily with SEO on fundamentals, and diverges on structure, proof, and freshness. See [SEO and GEO concepts](/seo-and-geo-concepts) and the [GEO checklist](/geo-checklist). ### Grain What one row of data represents: an order, a week, a channel. Recorded per metric in the [data dictionary](#data-dictionary) so two people summing the same numbers get the same answer. ### Guardrail metric A metric watched during an experiment to catch damage that the primary metric would hide. Related to a [counter-metric](#counter-metric), which does the same job in ongoing reporting. See [experiment brief](/experiment-brief). ### Hero article One of a capped set of at most 20 articles, one per [authority area](#authority-area), that a team commits to keeping current indefinitely. Chosen by what we can answer better than anyone, not by search volume. Distinct from a [supporting article](#supporting-article). See the [hero article register](/hero-article-register). ### Hook The opening line, subject, or first moment of a piece of copy, written to earn the reader's attention before anything else gets read. See [copy brief concepts](/copy-brief-concepts#sections-that-earn-their-place-conditionally). ### KPI Key performance indicator: a specific, named metric a campaign or test is measured against, defined once so everyone reports it the same way (for example, CPA or ROAS). "Metric" is the general word for any measured number; "KPI" is reserved for the metrics defined in [KPI definitions](/kpi-definitions). ### Lockup A fixed, pre-approved combination of logo elements, such as the mark plus a tagline or a partner's mark. Supplied as a single file and never rebuilt by hand. See [logo usage](/logo-usage). ### MER Marketing efficiency ratio: total revenue divided by total marketing spend. A blended, business-level metric rather than a per-channel one, which is what lets it sidestep attribution disputes entirely. See the [ecommerce metrics starter pack](/ecommerce-metrics-starter-pack). ### Message pillar One of the small, capped set of supporting arguments beneath a [value proposition](#value-proposition), each with its own proof. Capped so the document keeps rejecting things. See [messaging](/messaging). ### Meta description The summary shown under a page's title in a results list, and sometimes reused by an answer engine as a ready-made summary. Not a ranking factor, but the sentence that decides a click. Written after the article exists, not before. See [article standards](/article-standards). ### MQL Marketing qualified lead: a lead that meets an agreed scoring threshold for sales follow-up. The threshold should be agreed with sales in writing, not set by marketing alone. See the [sales-led metrics starter pack](/sales-led-metrics-starter-pack) and [SQL](#sql). ### Name jamming Packing multiple identifiers into a name using their short form instead of their long form, to stay within a platform's character limit. See [naming conventions overview](/naming-conventions-overview#name-jamming-short-identifiers). ### Naming convention The rules that define how campaign, ad set/ad group, and ad names are built from a shared dictionary of identifiers, so names stay predictable and filterable at scale. Start at the [naming conventions overview](/naming-conventions-overview). ### North star metric The single metric a team optimizes for when two reasonable options conflict. Not the most important [KPI](#kpi); the tiebreaker. There can be many KPIs and exactly one north star. See [reporting framework concepts](/reporting-framework-concepts#the-north-star-metric) and the [data dictionary](/marketing-reporting-data-dictionary#north-star-metric). ### Owner The one person accountable for a piece of work from start to finish. The owner drives it forward, makes the calls it needs, coordinates anyone else involved, and answers for the result. Ownership is singular: one named person, not a team or a role, so there's never a question of who to ask. ### Pacing Tracking spend against a budget over time to confirm a campaign will land on-target by the end of its flight, and adjusting before it over- or under-spends. See [budget pacing](/budget-pacing). ### Peeking Checking a running test repeatedly and stopping it the moment the result looks significant. Each extra look gives the result another chance to cross the threshold by luck, so stopping on a good day inflates the false-positive rate: the test reports a winner that isn't real. Avoid it by setting the sample size and end date before launch and letting the test run to that plan, whatever the numbers do along the way. Extending is only valid for an external reason (a traffic drop, a tracking break, a seasonal shift), never because the result hasn't reached significance yet. ### Point of difference What a brand does that its alternatives don't, that the target actually cares about. Only a difference if a competitor would dispute it. See [positioning](/positioning). ### Positioning statement A single working sentence naming the target, the [competitive frame](#competitive-frame), the [point of difference](#point-of-difference), and the [reason to believe](#reason-to-believe). Never used as copy. Used to test whether a campaign or headline fits. See [positioning](/positioning). ### Reason to believe The evidence that makes a [point of difference](#point-of-difference) credible: a number, a mechanism, or a named customer. The part that fails quietly when a claim outlives its source. See [positioning](/positioning). ### ROAS Return on ad spend: revenue attributed to a channel divided by that channel's ad spend. A per-channel, attributed metric, unlike the blended [MER](#mer). See the [ecommerce metrics starter pack](/ecommerce-metrics-starter-pack). ### Rollup Building a longer report from the shorter reports underneath it by summing raw inputs across the shorter periods and recalculating every ratio from the totals, rather than averaging the shorter periods' ratios directly. See [data quality and rollup concepts](/data-quality-and-rollup-concepts). ### Schema markup Machine-readable statements about what a page is and what it contains, sitting alongside the visible content. `Organization` markup with `sameAs` links is what connects separate mentions of a company into one recognized [entity](#entity). See [article standards](/article-standards). ### Search intent What someone is actually trying to accomplish with a query, as opposed to the words they typed. A full question typed into a chat window usually reveals it more clearly than three words typed into a search box. See [article brief concepts](/article-brief-concepts). ### Slug The part of a URL that identifies a specific page, after the domain and any directory. Lowercase, hyphenated, and without a year, so an article can be updated in place rather than republished at a new address. See [article standards](/article-standards). ### SOP Standard operating procedure: a documented, repeatable process for a recurring task, kept in the [Processes](/campaign-brief-process) tab. ### Source of truth The one approved system for a given metric. When two systems disagree, the approved one wins. See [reporting framework concepts](/reporting-framework-concepts#source-of-truth) and the [data dictionary](/marketing-reporting-data-dictionary#source-of-truth). ### SQL Sales qualified lead: a lead vetted by sales as ready for a direct conversation. Distinct from an [MQL](#mql), which is a marketing judgment rather than a sales one. See the [sales-led metrics starter pack](/sales-led-metrics-starter-pack). ### Standard identifier A permanent, always-present piece of a name (channel, objective, customer profile, offer, and so on) as opposed to a [testing identifier](#testing-identifier). See [naming conventions overview](/naming-conventions-overview#how-a-name-is-built). ### Statistical significance The confidence level that a test result reflects a real difference rather than random variation. See the full [statistical significance](/statistical-significance) guide. ### Supporting article Any article outside the [hero](#hero-article) set: how-tos, comparisons, definitions, and news pieces that build the cluster around a hero article. Reviewed less often and retired freely, but still owned and still tracked. See the [supporting article register](/supporting-article-register). ### Tagline A 3 to 8 word line that reinforces the position. Judged on memorability, where a [value proposition](#value-proposition) is judged on clarity. See [messaging](/messaging). ### Testing identifier An identifier appended to a name only while a hypothesis is actively being tested, marking which campaigns, ad sets, or ads belong to that test. See [naming conventions overview](/naming-conventions-overview#how-a-name-is-built) and the [experiment ID SOP](/experiment-ids). ### Tier 1 / 2 / 3 The three KPI levels: outcomes the business is directly accountable for, the drivers that move them, and the diagnostics that explain why a driver moved. See [reporting framework concepts](/reporting-framework-concepts#kpi-tiers). ### Topical authority Depth across a subject, built by covering a question, its sub-questions, and its edges in a connected cluster of articles rather than a single page. Distinct from entity authority, which is about being identifiable as a consistent [entity](#entity). See [SEO and GEO concepts](/seo-and-geo-concepts). ### Type scale The fixed set of sizes, line heights, and weights a brand's text uses, so hierarchy reads as hierarchy instead of as arbitrary sizes. See [visual identity](/visual-identity). ### UTM parameter A tag appended to a URL that records where a click came from, so traffic and conversions can be attributed back to a campaign, source, and medium. Use "UTM parameter," not "UTM tag." See [UTM parameters](/utm-parameters) and [UTM standards](/utm-standards). ### Value proposition One sentence, in the reader's language, saying what the brand does, for whom, and why it beats the alternative. The [positioning statement](#positioning-statement) rewritten for someone outside the company. See [messaging](/messaging). ### Voice trait A named characteristic of how a brand sounds, paired with a contrast that says where it stops ("direct, not blunt"). Without the contrast a trait stretches to justify anything. See [brand voice and tone](/brand-voice-and-tone). ## Related resources * [**Get started**](/overview) The setup path these terms describe. * [**Naming conventions overview**](/naming-conventions-overview) The conventions many of these terms belong to. * [**Data dictionary**](/marketing-reporting-data-dictionary) Full definitions, formulas and owners for every metric. * [**Brand guidelines**](/brand-guidelines) The index to the brand vocabulary. * [**New member onboarding**](/new-member-onboarding) Where a new reader starts.