system: LAKA SEO Grammar System version: '1.0' rule_count: 144 precedence: - safety_law_privacy_policy - user_value_and_accuracy - technical_eligibility_and_canonical_correctness - business_value_and_constraints - intent_fit_and_information_quality - measurement_and_learning - efficiency_and_scale - cosmetic_optimization rules: - id: BIZ-001 area: Business Value title: Define value before volume if: the program has no defined qualified business event then: define the event, qualification rule, owner, and estimated contribution value before optimizing traffic else: use the existing qualified event as the primary business outcome operators: - IF - THEN - ELSE - AND laka_levels: - Baseline primary_metric: measurement completeness guardrails: Do not substitute sessions or ranking counts for value. rationale: '' - id: BIZ-002 area: Business Value title: Traffic is diagnostic, not the north star if: organic traffic increases AND qualified conversions or margin do not then: diagnose intent quality, audience fit, next-action continuity, and attribution else: continue scaling only while downstream value and guardrails remain healthy operators: - IF - AND - THEN - ELSE laka_levels: - Baseline - Major primary_metric: incremental organic contribution margin guardrails: lead quality, sales capacity, refund rate rationale: '' - id: BIZ-003 area: Business Value title: Separate forecasts from actuals if: search volume, CTR, conversion, or revenue is estimated then: label the output as a forecast and record every assumption else: use observed query, conversion, revenue, and margin data operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: forecast error guardrails: confidence interval, source age rationale: '' - id: BIZ-004 area: Business Value title: Respect fulfillment capacity if: expected qualified demand exceeds delivery or sales capacity then: prioritize higher-margin, higher-fit clusters OR increase capacity before expanding volume else: optimize for profitable demand growth operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: contribution margin per capacity unit guardrails: response time, customer satisfaction rationale: '' - id: BIZ-005 area: Business Value title: Model offers separately if: different offers have materially different audiences, margins, sales cycles, or qualification rules then: assign separate value models and cluster portfolios else: a shared model may be used with explicit segment dimensions operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: value by offer guardrails: cross-offer cannibalization rationale: '' - id: BIZ-006 area: Business Value title: Count assisted value carefully if: an informational asset rarely receives last-click conversions but appears in qualified journeys then: measure assisted progression, returning users, branded demand, and controlled lift else: judge the asset on direct value and task completion operators: - IF - AND - THEN - ELSE laka_levels: - Baseline - Major primary_metric: assisted contribution value guardrails: double counting rationale: '' - id: BIZ-007 area: Business Value title: Split branded and non-branded demand if: brand familiarity materially changes intent, CTR, or conversion then: report branded and non-branded performance separately else: aggregate only for top-level totals operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: value-weighted non-brand clicks guardrails: brand campaign effects rationale: '' - id: BIZ-008 area: Business Value title: Include retention and expansion if: organic search serves existing customers, support, renewal, or repeat purchase tasks then: assign retention events and lifetime-value effects else: use acquisition outcomes for new-user clusters operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: organic-assisted retained margin guardrails: support deflection quality, churn rationale: '' - id: BIZ-009 area: Business Value title: Raise evidence thresholds for high stakes if: the topic can materially affect health, money, safety, legal rights, or major decisions then: require qualified review, primary sources, explicit limitations, and stronger maintenance else: apply the normal evidence standard operators: - IF - OR - THEN - ELSE - AND laka_levels: - Major - Structural primary_metric: verified task success guardrails: harm, complaint, correction rate rationale: '' - id: BIZ-010 area: Business Value title: Resolve competing goals explicitly if: traffic, lead quality, margin, brand, retention, and learning objectives conflict then: name one primary objective and classify the others as constraints or secondary outcomes else: use the single aligned objective operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: objective clarity guardrails: unintended metric trade-offs rationale: '' - id: BIZ-011 area: Business Value title: Allocate by marginal value if: multiple valid initiatives compete for limited resources then: rank them by expected incremental value divided by effort, risk, maintenance, and time to learning else: execute the only valid initiative operators: - IF - THEN - ELSE - AND laka_levels: - Baseline - Structural primary_metric: realized value per implementation hour guardrails: portfolio concentration risk rationale: '' - id: BIZ-012 area: Business Value title: Use economic stop rules if: expected remaining incremental value is below expected cost OR risk exceeds tolerance then: stop, defer, retire, or redesign the initiative else: continue to the next defined review point operators: - IF - OR - THEN - ELSE laka_levels: - Baseline - Paradigm primary_metric: net incremental value guardrails: sunk-cost bias rationale: '' - id: AUD-001 area: Audience and Journey title: Require an audience-task pair if: a candidate query cannot be mapped to a plausible audience and task then: classify it as uncertain and gather evidence else: attach the audience, condition, task, and journey state operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: classified-demand share guardrails: forced persona assumptions rationale: '' - id: AUD-002 area: Audience and Journey title: Treat funnel modifiers as heuristics if: a word such as how, best, review, price, or buy suggests a journey stage then: use it as a hypothesis and verify the actual task and result set else: infer stage from the full query, context, and behavior operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: intent-classification accuracy guardrails: stage stereotyping rationale: '' - id: AUD-003 area: Audience and Journey title: Model the funnel as a graph if: journey analysis assumes every user enters at awareness and moves linearly then: allow entry, exit, return, skip, and backward transitions else: retain observed journey paths operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: state-transition rate guardrails: attribution inflation rationale: '' - id: AUD-004 area: Audience and Journey title: Use adjacent next actions if: the proposed CTA requires substantially more trust or commitment than the current task supports then: offer an adjacent useful action OR add the proof needed to justify the transition else: measure the direct transition operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Major primary_metric: next-state transition rate guardrails: bounce, complaint, unsubscribe rationale: '' - id: AUD-005 area: Audience and Journey title: Split materially different audience needs if: two audience segments require different evidence, terminology, constraints, or offers then: create distinct sections, experiences, or canonical assets based on task compatibility else: serve them in one asset operators: - IF - OR - THEN - ELSE laka_levels: - Major - Structural primary_metric: qualified conversion by segment guardrails: duplication, maintenance rationale: '' - id: AUD-006 area: Audience and Journey title: Include accessibility conditions if: the task depends on visual, motor, auditory, cognitive, or assistive-technology access then: design the format, navigation, media, and controls for those conditions else: apply baseline accessible design to all assets operators: - IF - OR - THEN - ELSE - AND laka_levels: - Major - Structural primary_metric: task completion by accessibility condition guardrails: keyboard traps, missing alternatives rationale: '' - id: AUD-007 area: Audience and Journey title: Require real local distinctness if: a geography modifier is proposed as a separate page then: require actual service eligibility, local evidence, local differences, and maintainable data else: consolidate into a broader service-area asset operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: qualified local conversions guardrails: doorway-page risk, inaccurate claims rationale: '' - id: AUD-008 area: Audience and Journey title: Localize, do not merely translate if: language or country versions differ in terminology, regulation, inventory, currency, or user expectations then: localize content and targeting signals else: use a shared translation only where the task and facts truly remain equivalent operators: - IF - OR - THEN - ELSE laka_levels: - Major - Structural primary_metric: correct-locale conversion guardrails: translation quality, hreflang errors rationale: '' - id: AUD-009 area: Audience and Journey title: Account for device context if: device changes task conditions or interface needs then: adapt layout, media, interaction, and next action while preserving canonical meaning else: use the common responsive experience operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Major primary_metric: task completion by device guardrails: content parity rationale: '' - id: AUD-010 area: Audience and Journey title: Match evidence to urgency if: the user has high urgency or high perceived risk then: front-load direct answers, scope, availability, proof, and safe next actions else: use the standard information order operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: time to task completion guardrails: misleading urgency rationale: '' - id: AUD-011 area: Audience and Journey title: Explicitly exclude wrong audiences if: queries attract jobs, students, freebie seekers, unsupported geographies, or unrelated senses then: add exclusions to research, paid data, reporting, and content positioning else: retain the query in its valid cluster operators: - IF - OR - THEN - ELSE laka_levels: - Baseline - Minor primary_metric: irrelevant-impression and unqualified-lead rate guardrails: accidental exclusion of valuable demand rationale: '' - id: AUD-012 area: Audience and Journey title: Build post-purchase search pathways if: customers search for setup, use, troubleshooting, maintenance, renewal, or complementary products then: create support and retention task clusters with appropriate next actions else: focus the cluster on acquisition operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: retention and support value guardrails: support accuracy rationale: '' - id: SEM-001 area: Semantic Demand title: Start from parent concepts if: an offer lacks a mapped semantic territory then: identify parent concepts, neighboring categories, and business boundaries else: extend the existing concept graph operators: - IF - THEN - ELSE - AND laka_levels: - Baseline primary_metric: concept-map coverage guardrails: scope drift rationale: '' - id: SEM-002 area: Semantic Demand title: Expand through subtypes if: a parent concept contains meaningful narrower types with distinct tasks then: create subtype clusters and evaluate them independently else: keep subtype terms as sections or variants operators: - IF - AND - THEN - ELSE laka_levels: - Baseline - Major primary_metric: subtype demand coverage guardrails: thin fragmentation rationale: '' - id: SEM-003 area: Semantic Demand title: Normalize synonyms and paraphrases if: different phrases express the same sense and dominant task then: map them to one query cluster and one intended canonical asset else: investigate whether the task differs operators: - IF - AND - THEN - ELSE - XOR laka_levels: - Baseline primary_metric: cluster cohesion guardrails: missed senses rationale: '' - id: SEM-004 area: Semantic Demand title: Disambiguate homonyms and polysemy if: the same wording can refer to different concepts, entities, or tasks then: create sense labels and exclusions before clustering else: use the unambiguous concept operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: wrong-sense rate guardrails: query misclassification rationale: '' - id: SEM-005 area: Semantic Demand title: Model modifiers as facets if: terms vary by price, material, size, audience, platform, location, time, or other attributes then: represent them as attribute-value pairs and test whether they create a distinct task else: treat them as lexical variants operators: - IF - OR - THEN - ELSE laka_levels: - Baseline - Structural primary_metric: facet value and distinctness guardrails: facet explosion rationale: '' - id: SEM-006 area: Semantic Demand title: Use co-occurrence as context if: a term commonly appears near the topic but is not itself the user’s target then: use it only where it explains an entity, relation, attribute, or evidence need else: exclude it from mandatory coverage operators: - IF - AND - THEN - ELSE laka_levels: - Minor - Major primary_metric: task completeness guardrails: unnatural term insertion rationale: '' - id: SEM-007 area: Semantic Demand title: Map entities and relations if: the task depends on named things and how they connect then: record entity types and typed relations, then express those relations visibly else: use concept-only structure operators: - IF - AND - THEN - ELSE laka_levels: - Major primary_metric: entity-relation coverage guardrails: entity ambiguity rationale: '' - id: SEM-008 area: Semantic Demand title: Map query fan-out without page fan-out if: one broad task implies several supporting questions then: map the questions as sections, evidence needs, or distinct adjacent tasks else: do not create a separate page for every wording variation operators: - IF - THEN - ELSE - OR laka_levels: - Major - Structural primary_metric: fan-out task coverage guardrails: scaled-content duplication rationale: '' - id: SEM-009 area: Semantic Demand title: Triangulate demand evidence if: a candidate appears in only one tool or source then: seek corroboration from first-party data, users, result sets, competitors, or sales/support evidence else: increase confidence when independent sources agree operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: demand-confidence score guardrails: tool bias rationale: '' - id: SEM-010 area: Semantic Demand title: Do not equate zero tool volume with zero demand if: a query has low or zero reported volume but appears in first-party or customer evidence then: retain it as a low-volume or emerging candidate and measure the cluster else: deprioritize only after value and evidence review operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: observed cluster demand guardrails: anecdotal overreach rationale: '' - id: SEM-011 area: Semantic Demand title: Attach time state if: facts, prices, availability, standards, laws, events, or recommendations can change then: assign freshness requirements and update triggers else: classify the asset as stable evergreen operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: freshness compliance guardrails: stale claims rationale: '' - id: SEM-012 area: Semantic Demand title: Model emerging demand separately if: a new technology, regulation, event, or behavior has weak historical search data but strong causal evidence then: create an experimental demand thesis with explicit uncertainty else: use normal historical prioritization operators: - IF - AND - THEN - ELSE laka_levels: - Paradigm primary_metric: learning-adjusted emerging-demand value guardrails: trend chasing rationale: '' - id: CLU-001 area: Clustering and Canonical Ownership title: One dominant intent, one owner if: a dominant intent cluster is approved then: assign exactly one intended canonical URL else: keep the cluster in research operators: - IF - THEN - ELSE - EXACTLY_ONE laka_levels: - Baseline - Structural primary_metric: canonical ownership completeness guardrails: multiple owners rationale: '' - id: CLU-002 area: Clustering and Canonical Ownership title: Merge same-task candidates if: sense, task, answer structure, result class, and next action are compatible then: merge candidates into one cluster else: evaluate split conditions operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: demand coverage per canonical asset guardrails: over-broad pages rationale: '' - id: CLU-003 area: Clustering and Canonical Ownership title: Split different tasks if: queries share vocabulary but require different completion conditions or document forms then: create separate clusters and page roles else: keep them together operators: - IF - AND - OR - THEN - ELSE laka_levels: - Baseline - Major primary_metric: intent purity guardrails: duplicated introductions rationale: '' - id: CLU-004 area: Clustering and Canonical Ownership title: Use result overlap as evidence, not law if: current top results overlap strongly for two queries then: increase merge confidence but still inspect task and business context else: do not force a split solely because result overlap is low operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: cluster validation accuracy guardrails: SERP volatility rationale: '' - id: CLU-005 area: Clustering and Canonical Ownership title: Detect cannibalization by behavior if: multiple owned URLs receive the same cluster’s impressions with unstable dominance or divided value then: inspect intent, links, conversion, and canonical ownership else: allow multiple URLs when they serve genuinely different tasks operators: - IF - OR - THEN - ELSE laka_levels: - Baseline - Structural primary_metric: dominant URL share and URL entropy guardrails: false-positive cannibalization rationale: '' - id: CLU-006 area: Clustering and Canonical Ownership title: Merge with migration discipline if: two pages are redundant and one combined page is superior then: merge best information, select the winner, redirect the loser, and update links/sitemaps else: reposition or retain both operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: consolidated value and retained demand guardrails: migration loss rationale: '' - id: CLU-007 area: Clustering and Canonical Ownership title: Reposition a useful weaker page if: a weaker page has independent value but overlaps the stronger page’s task then: change its dominant intent and connect it as a supporting asset else: merge it operators: - IF - AND - THEN - ELSE laka_levels: - Major - Structural primary_metric: distinct-cluster value guardrails: residual overlap rationale: '' - id: CLU-008 area: Clustering and Canonical Ownership title: Inspect equity before deletion if: a page appears useless or redundant then: check links, traffic, conversions, references, and dependencies before delete/noindex else: retain or consolidate the page operators: - IF - OR - THEN - ELSE laka_levels: - Baseline - Structural primary_metric: value preserved after retirement guardrails: broken dependencies rationale: '' - id: CLU-009 area: Clustering and Canonical Ownership title: Align canonical signals if: duplicate or parameter URLs must remain accessible then: align canonical annotations, internal links, sitemap entries, redirects, and content signals else: self-canonicalize the unique page operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: declared-selected canonical agreement guardrails: mixed signals rationale: '' - id: CLU-010 area: Clustering and Canonical Ownership title: Protect established URLs if: an older URL has links, traffic, or history and the only proposed improvement is a prettier slug then: keep the URL unless expected benefit clearly exceeds migration risk else: optimize the new URL before launch operators: - IF - AND - THEN - ELSE laka_levels: - Minor - Structural primary_metric: post-change value retention guardrails: redirect chains rationale: '' - id: CLU-011 area: Clustering and Canonical Ownership title: Keep lexical variants together if: differences are spelling, word order, pluralization, abbreviation, or stemming without task change then: include natural variants in one asset and reporting cluster else: treat materially different senses separately operators: - IF - OR - THEN - ELSE laka_levels: - Minor primary_metric: variant coverage guardrails: keyword stuffing rationale: '' - id: CLU-012 area: Clustering and Canonical Ownership title: Test uncertain cluster boundaries if: merge/split evidence is inconclusive then: assign a provisional owner, record uncertainty, and observe query-to-URL behavior before scaling else: finalize the cluster operators: - IF - THEN - ELSE - UNTIL laka_levels: - Baseline - Major primary_metric: cluster stability guardrails: premature page multiplication rationale: '' - id: CON-001 area: Content, Format, and Evidence title: Select format from task if: the dominant task is known then: 'choose a format that completes that task: guide, table, calculator, map, diagnostic, template, video, or transaction page' else: research the task before drafting operators: - IF - THEN - ELSE - OR laka_levels: - Major primary_metric: task completion rate guardrails: format novelty without utility rationale: '' - id: CON-002 area: Content, Format, and Evidence title: Answer early if: a user can benefit from a concise direct answer then: provide it near the beginning and then support it else: state the path to the answer and necessary prerequisites operators: - IF - THEN - ELSE laka_levels: - Minor - Major primary_metric: time to first useful answer guardrails: oversimplification rationale: '' - id: CON-003 area: Content, Format, and Evidence title: Require information advantage if: the proposed asset only restates widely available material then: add original experience, data, testing, utility, synthesis, local proof, or decision support else: proceed with the defensible advantage operators: - IF - OR - THEN - ELSE laka_levels: - Major - Paradigm primary_metric: information-advantage score guardrails: unsupported originality claims rationale: '' - id: CON-004 area: Content, Format, and Evidence title: Prefer firsthand evidence where relevant if: the task involves reviews, demonstrations, processes, products, or lived outcomes then: show what was actually tested, observed, used, or built else: use the best available primary evidence operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: firsthand-evidence coverage guardrails: conflicts of interest rationale: '' - id: CON-005 area: Content, Format, and Evidence title: Make evidence visible if: a claim depends on a source, test, credential, or dataset then: place the supporting evidence where the user can inspect it else: qualify the claim as opinion or uncertainty operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Major primary_metric: verifiable-claim rate guardrails: citation mismatch rationale: '' - id: CON-006 area: Content, Format, and Evidence title: State limitations if: evidence has uncertainty, sample limits, conflicts, exceptions, or changing conditions then: state limitations and applicability boundaries else: present the stable conclusion operators: - IF - OR - THEN - ELSE laka_levels: - Minor primary_metric: correction and dispute rate guardrails: loss of nuance rationale: '' - id: CON-007 area: Content, Format, and Evidence title: Avoid arbitrary length targets if: a word count is proposed without a task-based reason then: define completeness by required questions, evidence, and decision support else: use length as a production estimate only operators: - IF - THEN - ELSE - AND laka_levels: - Baseline - Major primary_metric: task completeness per reading time guardrails: content bloat rationale: '' - id: CON-008 area: Content, Format, and Evidence title: Govern AI-assisted content if: AI assists research, drafting, transformation, or scale then: require factual verification, original value, editorial ownership, and policy compliance else: use the same quality controls for human-only work operators: - IF - OR - THEN - ELSE - AND laka_levels: - Structural primary_metric: verified publication quality guardrails: hallucination, scaled abuse rationale: '' - id: CON-009 area: Content, Format, and Evidence title: Require row-level uniqueness at scale if: a template generates pages from a dataset then: require accurate unique data, distinct user utility, and a valid canonical task for each row else: consolidate rows into filters, tables, or broader assets operators: - IF - AND - THEN - ELSE laka_levels: - Structural - Paradigm primary_metric: unique-value coverage per generated page guardrails: doorway and thin-page risk rationale: '' - id: CON-010 area: Content, Format, and Evidence title: Use media when it improves the task if: an image, diagram, video, audio, or interactive can explain or demonstrate more effectively then: include accessible, relevant media and measure its use else: keep the experience text-first where sufficient operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: media-assisted task completion guardrails: performance, accessibility rationale: '' - id: CON-011 area: Content, Format, and Evidence title: Define refresh triggers if: content depends on changing data or conditions then: assign owner, source, review cadence, and event-triggered updates else: review only when performance or business context changes operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: freshness SLA guardrails: maintenance debt rationale: '' - id: CON-012 area: Content, Format, and Evidence title: Support commercial pages with useful assets if: a commercial page is important but difficult to cite or insufficient for research-stage users then: create genuinely useful supporting evidence, tools, comparisons, or guides else: improve the commercial page directly operators: - IF - AND - OR - THEN - ELSE laka_levels: - Major - Structural primary_metric: assisted commercial value guardrails: manufactured link bait rationale: '' - id: ONP-001 area: On-page and Search Appearance title: Align the title with the task if: the title does not clearly identify the page’s dominant task or value then: rewrite it truthfully for relevance and useful differentiation else: retain it and monitor operators: - IF - OR - THEN - ELSE laka_levels: - Minor primary_metric: position-adjusted CTR guardrails: qualified conversion rate rationale: '' - id: ONP-002 area: On-page and Search Appearance title: Use unique titles for unique canonicals if: two intended indexable canonical pages use materially duplicate titles then: differentiate titles according to their distinct tasks else: allow shared patterns with unique identifying content operators: - IF - AND - THEN - ELSE laka_levels: - Minor - Structural primary_metric: duplicate-title rate guardrails: template inconsistency rationale: '' - id: ONP-003 area: On-page and Search Appearance title: Measure title changes causally if: a title or snippet is changed then: record change date, exposed queries/pages, expected mechanism, and CTR guardrails else: do not claim impact operators: - IF - THEN - ELSE - AND laka_levels: - Minor primary_metric: CTR residual guardrails: ranking, conversion, seasonality rationale: '' - id: ONP-004 area: On-page and Search Appearance title: Write descriptions for decision support if: the result snippet can help users determine relevance then: write a concise truthful proposition with task, benefit, and distinguishing detail else: allow the search engine to select contextual text operators: - IF - AND - THEN - ELSE laka_levels: - Minor primary_metric: organic CTR guardrails: snippet mismatch rationale: '' - id: ONP-005 area: On-page and Search Appearance title: Use headings as answer architecture if: the task has stages, criteria, subquestions, or choices then: structure headings around those information functions else: use a simple hierarchy operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: section-level task coverage guardrails: heading keyword stuffing rationale: '' - id: ONP-006 area: On-page and Search Appearance title: Write natural body language if: required concepts, entities, attributes, and relations are missing then: add them where they improve the explanation else: do not insert terms solely for coverage scores operators: - IF - AND - THEN - ELSE laka_levels: - Major primary_metric: semantic task completeness guardrails: readability rationale: '' - id: ONP-007 area: On-page and Search Appearance title: Reject keyword stuffing if: repetition exists primarily to influence ranking rather than aid comprehension then: remove or rewrite it and preserve only natural references else: retain meaningful repetition operators: - IF - THEN - ELSE laka_levels: - Minor primary_metric: editorial quality guardrails: loss of clarity rationale: '' - id: ONP-008 area: On-page and Search Appearance title: Describe meaningful images if: an image communicates information or functions as a link/control then: provide appropriate descriptive alternative text else: use empty alternative text for purely decorative images operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Structural primary_metric: accessible-media coverage guardrails: redundant alt text rationale: '' - id: ONP-009 area: On-page and Search Appearance title: Optimize new URLs, protect old ones if: a page has not launched then: use a concise descriptive stable URL else: avoid changing an established URL without a migration case operators: - IF - THEN - ELSE laka_levels: - Minor - Structural primary_metric: URL stability and migration retention guardrails: redirect chains rationale: '' - id: ONP-010 area: On-page and Search Appearance title: Use structured data only when applicable if: a supported structured-data type accurately represents visible page content then: implement and validate it else: do not add irrelevant or hidden markup operators: - IF - AND - THEN - ELSE laka_levels: - Major - Structural primary_metric: valid eligible items guardrails: manual action, markup drift rationale: '' - id: ONP-011 area: On-page and Search Appearance title: Measure rich-result realization if: structured data validates successfully then: measure actual search appearance, impressions, CTR, and value else: debug eligibility or accept that display is not guaranteed operators: - IF - THEN - ELSE laka_levels: - Baseline - Major primary_metric: rich-result appearance and value guardrails: misattribution rationale: '' - id: ONP-012 area: On-page and Search Appearance title: Adjust CTR expectations if: CTR is evaluated then: control for position, device, country, brand status, query class, and search appearance else: do not use a universal CTR threshold operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: CTR residual guardrails: low sample size rationale: '' - id: GRF-001 area: Internal Graph and Architecture title: Link every priority page if: a page is intended to receive organic visibility then: ensure at least one crawlable contextual internal link points to it else: remove it from the priority/index set if it has no valid role operators: - IF - THEN - ELSE - AT_LEAST laka_levels: - Baseline - Minor primary_metric: priority-page orphan rate guardrails: forced links rationale: '' - id: GRF-002 area: Internal Graph and Architecture title: Use descriptive anchors if: anchor text is generic and a natural descriptive label is available then: replace it with concise context-setting text else: retain the generic control only when interface convention requires it operators: - IF - AND - THEN - ELSE laka_levels: - Minor primary_metric: descriptive-anchor coverage guardrails: over-optimization rationale: '' - id: GRF-003 area: Internal Graph and Architecture title: Require user context for links if: a proposed link does not help the reader understand, compare, verify, or take the next step then: do not add it else: add it in the relevant sentence or interface operators: - IF - OR - THEN - ELSE laka_levels: - Minor primary_metric: useful internal-link interaction guardrails: link clutter rationale: '' - id: GRF-004 area: Internal Graph and Architecture title: Create hubs for navigation needs if: a coherent task family requires users to choose among multiple paths then: create or improve a hub with clear categories and routes else: connect assets contextually without a dedicated hub operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: hub-assisted discovery and conversion guardrails: empty taxonomy pages rationale: '' - id: GRF-005 area: Internal Graph and Architecture title: Validate supporting assets if: an adjacent question has independent demand, a complete answer, and a useful relationship to a core asset then: create a supporting asset and connect it else: include it as a section or FAQ operators: - IF - AND - THEN - ELSE laka_levels: - Major primary_metric: supporting-asset assisted value guardrails: thin content rationale: '' - id: GRF-006 area: Internal Graph and Architecture title: Reduce excessive depth if: important pages are difficult to reach through internal navigation then: add meaningful paths, hubs, or template links else: retain depth for low-value or intentionally isolated pages operators: - IF - THEN - ELSE - OR laka_levels: - Structural primary_metric: priority-page click depth guardrails: navigation overload rationale: '' - id: GRF-007 area: Internal Graph and Architecture title: Use strong internal sources deliberately if: an authoritative owned page can naturally support a high-value target then: add a contextual link where it benefits the reader else: do not insert an unrelated link solely to move authority operators: - IF - AND - THEN - ELSE laka_levels: - Minor primary_metric: target visibility lift guardrails: source-page task disruption rationale: '' - id: GRF-008 area: Internal Graph and Architecture title: Keep topical communities coherent if: a link crosses unrelated topic communities without user need then: remove or avoid it else: retain cross-topic links that represent a real prerequisite, comparison, or journey edge operators: - IF - AND - NOT - THEN - ELSE laka_levels: - Structural primary_metric: topical edge precision guardrails: over-siloing rationale: '' - id: GRF-009 area: Internal Graph and Architecture title: Make links crawlable if: navigation or internal discovery relies on nonstandard scripted interactions then: expose crawlable anchors with destinations and validate rendered HTML else: retain standard crawlable links operators: - IF - THEN - ELSE - AND laka_levels: - Structural primary_metric: crawlable-link coverage guardrails: duplicate navigation rationale: '' - id: GRF-010 area: Internal Graph and Architecture title: Concentrate, do not hoard if: many low-value pages dilute attention and internal graph flow then: consolidate, noindex, archive, or demote according to user and business value else: retain useful nodes operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: value-weighted index and graph concentration guardrails: loss of long-tail value rationale: '' - id: GRF-011 area: Internal Graph and Architecture title: Update the graph after migrations if: a page is merged, redirected, renamed, or retired then: update internal links, navigation, canonicals, and sitemaps to the final destination else: leave the graph unchanged operators: - IF - OR - THEN - ELSE - AND laka_levels: - Structural primary_metric: direct-to-final-link rate guardrails: redirect loops rationale: '' - id: GRF-012 area: Internal Graph and Architecture title: Measure paths, not link counts alone if: internal linking is evaluated then: measure discovery, depth, target visibility, task progression, and conversion paths else: use raw link counts only as diagnostics operators: - IF - THEN - ELSE - AND laka_levels: - Baseline primary_metric: graph-assisted value guardrails: correlation mistaken for causation rationale: '' - id: AUT-001 area: External Authority and Earned Evidence title: Evaluate link quality multidimensionally if: an external link opportunity is considered then: evaluate relevance, editorial independence, source credibility, placement, uniqueness, and persistence else: reject or deprioritize the opportunity operators: - IF - AND - THEN - ELSE laka_levels: - Baseline primary_metric: qualified retained referring domains guardrails: spam and paid-link risk rationale: '' - id: AUT-002 area: External Authority and Earned Evidence title: Create a linkable reason if: outreach asks for attention without a concrete user benefit then: build or identify data, utility, evidence, correction, visualization, or unique reference value else: lead with the existing value operators: - IF - AND - NOT - THEN - ELSE - OR laka_levels: - Major - Paradigm primary_metric: placement rate guardrails: misleading claims rationale: '' - id: AUT-003 area: External Authority and Earned Evidence title: Support difficult commercial targets if: the target commercial page is unlikely to earn independent citations then: earn links to a relevant useful asset and connect it to the target OR improve the target’s reference value else: promote the target directly operators: - IF - OR - THEN - ELSE laka_levels: - Major - Structural primary_metric: assisted target-page visibility guardrails: manipulative intermediary pages rationale: '' - id: AUT-004 area: External Authority and Earned Evidence title: Reject manipulative placements if: a link requires deception, irrelevant insertion, undisclosed payment, or policy violation then: reject it else: proceed with transparent editorial or appropriately attributed placement operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: policy-compliant placement rate guardrails: manual actions, reputation rationale: '' - id: AUT-005 area: External Authority and Earned Evidence title: Prefer unique relevant domains if: many links come repeatedly from the same source while relevant independent sources are available then: diversify outreach toward unique relevant domains else: deepen the useful relationship without treating every link as independent authority operators: - IF - AND - THEN - ELSE laka_levels: - Major primary_metric: qualified unique referring domains guardrails: diversity for its own sake rationale: '' - id: AUT-006 area: External Authority and Earned Evidence title: Prefer contextual editorial placement if: a link can appear in a useful body context or only in a repeated footer/sidebar then: prefer the contextual editorial placement else: treat repeated placement as lower-confidence evidence operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: contextual placement share guardrails: sponsorship disclosure rationale: '' - id: AUT-007 area: External Authority and Earned Evidence title: Use tool metrics as proxies if: DA, DR, UR, PA, traffic estimates, or difficulty scores are used then: treat them as vendor-specific diagnostics and compare within one system else: use direct observations where available operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: proxy-to-outcome calibration guardrails: cross-tool comparisons rationale: '' - id: AUT-008 area: External Authority and Earned Evidence title: Investigate lost evidence if: a valuable relevant link or mention disappears then: classify cause and recover only when the source relationship and user value remain valid else: record normal churn operators: - IF - AND - THEN - ELSE laka_levels: - Minor - Major primary_metric: valuable-link recovery rate guardrails: harassing outreach rationale: '' - id: AUT-009 area: External Authority and Earned Evidence title: Measure brand demand effects if: earned media or authoritative mentions may increase awareness without direct referral clicks then: track branded queries, direct visits, assisted conversions, and target visibility else: measure direct referral and link effects operators: - IF - AND - NOT - THEN - ELSE laka_levels: - Baseline - Major primary_metric: incremental branded demand guardrails: campaign overlap rationale: '' - id: AUT-010 area: External Authority and Earned Evidence title: Segment outreach by relationship if: prospects differ by audience, reason to care, or editorial context then: create distinct outreach propositions and evidence packages else: use one aligned proposition operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: positive reply and placement rate by segment guardrails: automation quality rationale: '' - id: AUT-011 area: External Authority and Earned Evidence title: Compound original research if: original data or a benchmark performs well then: refresh it, expose methodology, create reusable views, and connect related clusters else: treat it as a one-time asset operators: - IF - THEN - ELSE - AND laka_levels: - Structural - Paradigm primary_metric: research-attributed links and value guardrails: methodology drift rationale: '' - id: AUT-012 area: External Authority and Earned Evidence title: Judge authority by target outcomes if: links or mentions increase but the target cluster does not improve then: inspect relevance, target selection, indexing, content fit, and time lag else: continue scaling the validated authority mechanism operators: - IF - AND - THEN - ELSE laka_levels: - Baseline - Major primary_metric: incremental target-cluster visibility and value guardrails: over-attribution rationale: '' - id: TEC-001 area: Technical Eligibility title: Require public accessibility if: a page is intended for search visibility then: ensure users and permitted crawlers can request its essential content else: exclude it intentionally operators: - IF - THEN - ELSE - AND laka_levels: - Baseline - Structural primary_metric: accessible-priority-URL rate guardrails: private or sensitive exposure rationale: '' - id: TEC-002 area: Technical Eligibility title: Use correct response states if: a canonical page is healthy and available then: return a successful response; use redirects, not-found, or server errors only for their true states else: repair the response logic operators: - IF - THEN - ELSE - OR laka_levels: - Structural primary_metric: correct-status rate guardrails: soft 404s rationale: '' - id: TEC-003 area: Technical Eligibility title: Validate rendered main content if: essential content or links depend on JavaScript then: inspect rendered output and ensure resources are accessible else: use the server-delivered content operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: rendered-content parity guardrails: hydration and blocked-resource errors rationale: '' - id: TEC-004 area: Technical Eligibility title: Separate crawling from indexing controls if: the goal is to keep a page out of search results then: use an appropriate noindex, authentication, removal, redirect, or deletion method else: use robots.txt only to manage crawler requests, not as a deindex guarantee operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: intended index-state accuracy guardrails: content leakage rationale: '' - id: TEC-005 area: Technical Eligibility title: Maintain canonical coherence if: multiple URLs represent substantially the same content then: select a preferred URL and align canonical, redirects where appropriate, links, and sitemap signals else: self-canonicalize the unique asset where useful operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: canonical agreement guardrails: canonical chains rationale: '' - id: TEC-006 area: Technical Eligibility title: Keep sitemaps canonical and current if: a URL is important, canonical, and intended for indexing then: include its clean URL and accurate modification state in the sitemap else: exclude noncanonical, redirected, error, or intentionally nonindexed URLs operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: valid canonical sitemap coverage guardrails: misleading lastmod rationale: '' - id: TEC-007 area: Technical Eligibility title: Redirect to the closest valid successor if: a URL permanently moves or is consolidated then: redirect directly to the most relevant final destination and update internal references else: serve the existing URL operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: direct permanent-redirect accuracy guardrails: chains, loops, irrelevant redirects rationale: '' - id: TEC-008 area: Technical Eligibility title: Treat not-found pages intentionally if: a resource no longer exists and has no valid replacement then: return a real not-found state with helpful navigation else: redirect only when a genuine equivalent exists operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: soft-404 rate guardrails: lost valuable links rationale: '' - id: TEC-009 area: Technical Eligibility title: Control faceted and parameter volume if: filters, sorting, search, tracking, or session parameters create many near-duplicate URLs then: define crawl, canonical, link, and index rules by facet value else: allow unique valuable combinations operators: - IF - OR - THEN - ELSE - AND laka_levels: - Structural primary_metric: valuable-crawl share guardrails: index explosion rationale: '' - id: TEC-010 area: Technical Eligibility title: Preserve content parity if: mobile, desktop, locale, or rendered versions differ then: ensure essential content, metadata, structured data, and links remain equivalent where they should else: document intentional differences operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: version parity guardrails: cloaking-like mismatches rationale: '' - id: TEC-011 area: Technical Eligibility title: Prioritize real-user performance if: field performance or business outcomes show latency, instability, or interaction problems then: fix the highest-exposure causes and measure Core Web Vitals plus conversion/task outcomes else: repair only obvious low-cost issues and avoid score chasing operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Structural primary_metric: good-CWV exposure and task value guardrails: functionality regression rationale: '' - id: TEC-012 area: Technical Eligibility title: Scale crawl work to site size and change rate if: the site is very large, rapidly changing, or shows crawl/index waste then: invest in crawl-path, logs, facets, and index-quality optimization else: use a lighter periodic technical audit operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: valuable indexed URLs per crawl/resource unit guardrails: technical work without business case rationale: '' - id: CVR-001 area: Conversion and Experience title: Complete the search task before selling if: the page asks for conversion before delivering the promised answer then: front-load useful task completion and place the offer at a logical transition else: retain the aligned conversion path operators: - IF - BEFORE - THEN - ELSE laka_levels: - Major primary_metric: task completion then conversion guardrails: reduced qualified action rationale: '' - id: CVR-002 area: Conversion and Experience title: Match nudge to journey state if: the next action is not adjacent to the user’s current state then: choose a lower-friction or more relevant action else: measure the existing transition operators: - IF - THEN - ELSE laka_levels: - Minor - Major primary_metric: next-state conversion rate guardrails: lead quality rationale: '' - id: CVR-003 area: Conversion and Experience title: Optimize speed for outcomes if: slow or unstable experience correlates with abandonment or conversion loss then: fix the responsible bottleneck and validate field/user outcomes else: do not pursue a perfect tool score solely for ranking operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Structural primary_metric: conversion and task completion by performance band guardrails: broken features rationale: '' - id: CVR-004 area: Conversion and Experience title: Reduce form friction selectively if: form abandonment is high AND removed fields are not needed for qualification or routing then: shorten, stage, or clarify the form else: retain fields and improve explanation operators: - IF - AND - THEN - ELSE - OR laka_levels: - Minor - Major primary_metric: qualified form completion guardrails: unqualified volume rationale: '' - id: CVR-005 area: Conversion and Experience title: Optimize for lead quality if: lead count rises while acceptance, close rate, or margin falls then: tighten intent, promise, qualification, or routing else: scale the validated acquisition path operators: - IF - AND - OR - THEN - ELSE laka_levels: - Major - Structural primary_metric: sales-accepted organic leads guardrails: total pipeline rationale: '' - id: CVR-006 area: Conversion and Experience title: Test CTA propositions, not colors alone if: a CTA underperforms then: test task relevance, value, risk, evidence, commitment, placement, and wording before cosmetic variables else: retain and scale it operators: - IF - THEN - ELSE - BEFORE laka_levels: - Minor - Major primary_metric: qualified CTA conversion guardrails: misleading claims rationale: '' - id: CVR-007 area: Conversion and Experience title: Place trust at the decision point if: users hesitate because scope, proof, identity, terms, price, or risk is unclear then: place relevant evidence and clarification near the decision else: avoid repetitive trust decoration operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: decision completion rate guardrails: privacy and disclosure rationale: '' - id: CVR-008 area: Conversion and Experience title: Avoid premature hard selling if: the query is early-stage and commercial pressure harms task satisfaction then: use a useful soft transition such as a related guide, tool, save, or diagnostic else: present the proportional commercial action operators: - IF - AND - THEN - ELSE - OR laka_levels: - Major primary_metric: assisted qualified progression guardrails: unsubscribe, bounce rationale: '' - id: CVR-009 area: Conversion and Experience title: Design retention paths if: the organic visitor is an existing customer or user then: route to successful use, support, renewal, expansion, or advocacy else: route prospects toward acquisition operators: - IF - OR - THEN - ELSE laka_levels: - Major primary_metric: retention/expansion event value guardrails: support escalation rationale: '' - id: CVR-010 area: Conversion and Experience title: Measure satisfaction with task-specific events if: generic engagement metrics cannot show whether the task was completed then: instrument meaningful events such as calculator completion, comparison use, download, copy, or successful route else: use the existing task event operators: - IF - THEN - ELSE - OR laka_levels: - Structural primary_metric: task completion rate guardrails: event inflation rationale: '' - id: CVR-011 area: Conversion and Experience title: Preserve message continuity if: the search promise, landing answer, CTA, and sales follow-up describe different outcomes then: align the chain around one user task and offer else: retain the coherent journey operators: - IF - AND - THEN - ELSE laka_levels: - Structural primary_metric: promise-to-close continuity guardrails: overly narrow messaging rationale: '' - id: CVR-012 area: Conversion and Experience title: Protect user and brand guardrails if: a conversion intervention increases action but also complaints, refunds, low-quality leads, or distrust then: revert, constrain, or redesign the intervention else: continue while guardrails stay within tolerance operators: - IF - AND - OR - THEN - ELSE laka_levels: - Minor - Major primary_metric: guardrail-adjusted conversion value guardrails: dark patterns rationale: '' - id: EXP-001 area: Measurement and Experimentation title: Require a mechanism hypothesis if: an intervention has no stated reason it should affect an outcome then: write an object → action → mechanism → outcome hypothesis else: execute the stated hypothesis operators: - IF - THEN - ELSE laka_levels: - Baseline primary_metric: hypothesis completeness guardrails: post-hoc storytelling rationale: '' - id: EXP-002 area: Measurement and Experimentation title: Capture a baseline if: a change is about to be implemented then: record pre-change performance, context, exposure, and data quality else: do not claim incremental effect operators: - IF - THEN - ELSE - AND laka_levels: - Baseline primary_metric: baseline completeness guardrails: seasonal distortion rationale: '' - id: EXP-003 area: Measurement and Experimentation title: Use a change ledger if: any SEO-relevant production change occurs then: assign a change ID, owner, scope, time, expected mechanism, and rollback state else: retain the previous ledger operators: - IF - THEN - ELSE - AND laka_levels: - Structural primary_metric: change attribution coverage guardrails: logging overhead rationale: '' - id: EXP-004 area: Measurement and Experimentation title: Choose one primary metric if: an experiment has many possible success metrics then: select one primary decision metric and label others secondary or guardrail else: use the aligned primary metric operators: - IF - THEN - ELSE - EXACTLY_ONE laka_levels: - Baseline primary_metric: decision consistency guardrails: metric cherry-picking rationale: '' - id: EXP-005 area: Measurement and Experimentation title: Set guardrails before launch if: a change can improve the primary metric while harming quality, accessibility, revenue, or trust then: define guardrail thresholds and rollback behavior else: use standard site guardrails operators: - IF - OR - THEN - ELSE - AND laka_levels: - Baseline primary_metric: guardrail breach rate guardrails: unobserved harm rationale: '' - id: EXP-006 area: Measurement and Experimentation title: Use controls when feasible if: comparable untreated pages, clusters, regions, or time periods exist then: use a control or difference-in-differences design else: use interrupted time series with explicit limitations operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: estimated incremental lift guardrails: control contamination rationale: '' - id: EXP-007 area: Measurement and Experimentation title: Account for seasonality and external events if: demand, algorithms, campaigns, news, inventory, or holidays can affect the outcome then: annotate and model those factors before attributing change else: use the stable comparison operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: context-adjusted lift guardrails: overfitting rationale: '' - id: EXP-008 area: Measurement and Experimentation title: Represent uncertainty if: sample size is small, variance is high, or exposure is incomplete then: report ranges, confidence, and directional evidence else: report the stable estimate operators: - IF - OR - THEN - ELSE laka_levels: - Baseline primary_metric: estimate uncertainty guardrails: false precision rationale: '' - id: EXP-009 area: Measurement and Experimentation title: Allow processing and learning time if: a change has not had enough crawl, index, ranking, or behavioral exposure then: continue observation until the predefined minimum exposure or stop condition else: make the decision operators: - IF - OR - THEN - ELSE - UNTIL laka_levels: - Baseline primary_metric: minimum-exposure completion guardrails: indefinite waiting rationale: '' - id: EXP-010 area: Measurement and Experimentation title: Distinguish single-variable and package tests if: several material variables change together then: classify the intervention as a package test OR separate variables for causal learning else: treat it as a focused test operators: - IF - OR - THEN - ELSE laka_levels: - Minor - Major - Structural primary_metric: attribution confidence guardrails: slow recovery rationale: '' - id: EXP-011 area: Measurement and Experimentation title: Escalate by diagnosed mechanism if: a lower-level LAKA intervention fails then: re-diagnose and move to Major, Structural, or Paradigm only when the failure mechanism requires it else: retain or scale the successful level operators: - IF - THEN - ELSE - OR laka_levels: - Minor - Major - Structural - Paradigm primary_metric: value gained per escalation guardrails: unnecessary redesign rationale: '' - id: EXP-012 area: Measurement and Experimentation title: Use explicit continue/revise/revert/retire decisions if: a review point is reached then: choose exactly one decision based on primary metric, guardrails, cost, and confidence else: continue observation only under a predefined uncertainty rule operators: - IF - AND - THEN - ELSE - XOR laka_levels: - Baseline primary_metric: decision closure rate guardrails: zombie experiments rationale: '' - id: GOV-001 area: Generative Search, Scale, and Governance title: Keep foundational SEO as the base if: the goal includes visibility in generative search experiences then: apply the same valuable-content, crawlability, indexability, and search-quality foundations else: apply normal search foundations operators: - IF - AND - THEN - ELSE laka_levels: - Baseline - Structural primary_metric: search and generative visibility value guardrails: channel-specific hacks rationale: '' - id: GOV-002 area: Generative Search, Scale, and Governance title: Do not depend on llms.txt for Google Search if: a proposal claims a special AI text file is required for Google visibility then: reject the requirement and prioritize public crawlable indexable content else: maintain such a file only for a separate system that explicitly uses it operators: - IF - THEN - ELSE laka_levels: - Baseline primary_metric: eligible indexed coverage guardrails: maintenance distraction rationale: '' - id: GOV-003 area: Generative Search, Scale, and Governance title: Do not chunk content mechanically if: content is being split only because an AI system is assumed to need tiny sections then: organize by human task, readability, and information structure else: retain useful natural sections operators: - IF - THEN - ELSE laka_levels: - Major primary_metric: task completion and retrieval coverage guardrails: fragmentation rationale: '' - id: GOV-004 area: Generative Search, Scale, and Governance title: Use query fan-out as a research map if: a broad request implies related information needs then: map prerequisite, comparison, evidence, and follow-up needs across the cluster graph else: use the direct task map operators: - IF - THEN - ELSE - AND laka_levels: - Major - Structural primary_metric: fan-out information coverage guardrails: one-page-per-query expansion rationale: '' - id: GOV-005 area: Generative Search, Scale, and Governance title: Block scaled content abuse if: content volume is generated primarily to manipulate rankings or AI responses without distinct user value then: stop publication and redesign the system else: continue only with row-level utility, accuracy, and governance operators: - IF - AND - NOT - THEN - ELSE laka_levels: - Structural - Paradigm primary_metric: unique useful asset rate guardrails: spam-policy exposure rationale: '' - id: GOV-006 area: Generative Search, Scale, and Governance title: Measure generative visibility directly if: Search Console or another first-party surface provides generative-search reporting then: track it as a distinct search appearance and connect it to downstream value else: use broader search and referral diagnostics with limitations operators: - IF - THEN - ELSE - AND laka_levels: - Baseline - Structural primary_metric: generative-search qualified value guardrails: reporting coverage rationale: '' - id: GOV-007 area: Generative Search, Scale, and Governance title: Do not invent AI-specific schema if: a proposal adds unsupported structured data solely for generative visibility then: reject it; use supported structured data only when it matches visible content else: validate the applicable markup operators: - IF - THEN - ELSE laka_levels: - Baseline - Structural primary_metric: valid supported markup guardrails: markup abuse rationale: '' - id: GOV-008 area: Generative Search, Scale, and Governance title: Prepare for agentic use through usability if: browser agents or assistive systems need to interpret and operate the site then: improve semantic controls, labels, accessible structure, clear states, and stable data else: retain the usable human-facing implementation operators: - IF - OR - THEN - ELSE - AND laka_levels: - Structural - Paradigm primary_metric: successful task completion by automated/assistive clients guardrails: security and consent rationale: '' - id: GOV-009 area: Generative Search, Scale, and Governance title: Expose freshness and provenance if: data may change or be reused in generated answers then: show dates, methodology, sources, ownership, and update state where useful else: omit unnecessary metadata operators: - IF - OR - THEN - ELSE - AND laka_levels: - Major primary_metric: freshness/provenance completeness guardrails: false recency rationale: '' - id: GOV-010 area: Generative Search, Scale, and Governance title: Constrain programmatic expansion if: candidate combinations exceed quality or maintenance capacity then: raise gates, consolidate, stage release, or automate verified data maintenance else: publish within capacity operators: - IF - OR - THEN - ELSE laka_levels: - Structural primary_metric: maintained useful-page rate guardrails: index and maintenance debt rationale: '' - id: GOV-011 area: Generative Search, Scale, and Governance title: Protect privacy, consent, and sensitive data if: volumetric generation or personalization uses personal, confidential, licensed, or restricted data then: minimize data, obtain rights/consent, secure access, and exclude unsafe outputs else: use public or owned permissible data operators: - IF - OR - THEN - ELSE - AND laka_levels: - Structural primary_metric: governance compliance guardrails: privacy harm rationale: '' - id: GOV-012 area: Generative Search, Scale, and Governance title: Pursue paradigm assets selectively if: a new tool, dataset, benchmark, diagnostic, or agent can complete the task far better than existing result classes then: run a bounded paradigm experiment with a defensibility and maintenance plan else: improve the current asset class operators: - IF - OR - THEN - ELSE - AND laka_levels: - Paradigm primary_metric: new-demand and incremental value creation guardrails: build cost, adoption risk rationale: ''