Bow Tie Kreative SEO System
Evaluate a candidate

Operating

11 phases, each with an exit gate

A dependency-ordered rollout. Measurement and technical eligibility come before expansion, because an unmeasurable or ineligible asset cannot pay back what it costs.

Phase 0 Governance and workspace
  • project owner;
  • site and offer scope;
  • source-of-truth repository;
  • change ledger;
  • access and privacy rules;
  • naming conventions;
  • rule-library version.
Exit gate
owner
AND scope
AND permissions
AND change control
are defined.
Phase 1 Business and measurement baseline
  • Define offers and audiences.
  • Define qualified value events.
  • Record margin or value bands.
  • verify Search Console and analytics.
  • connect CRM/commerce outcomes where feasible.
  • classify brand/non-brand.
  • create executive dashboard.
  • test the event chain.
  • document limitations.
  • capture baseline.
Phase 2 Technical eligibility baseline
  • Inventory intended canonical URLs.
  • classify intended index state.
  • inspect status, robots, rendering, canonical, index, sitemap.
  • calculate eligible canonical coverage.
  • calculate canonical agreement.
  • identify orphan and depth issues.
  • inspect duplicate/facet volume.
  • inspect field performance by exposure.
  • rank issues by affected business value.
  • repair P0 blockers.
Phase 3 Existing opportunity recovery
  • join queries, URLs, conversions, and value.
  • cluster current queries.
  • find strong pages with weak targeting.
  • find high-impression pages with low CTR residual.
  • find converting pages with weak visibility.
  • find ranking pages with poor qualification.
  • find cannibalizing clusters.
  • find pages with authority but no useful target.
  • score opportunities.
  • launch the first bounded Minor/Major cohort.
Phase 4 Semantic and canonical architecture
  • Build audience-task map.
  • build semantic graph.
  • collect query and customer language.
  • normalize synonyms and variants.
  • disambiguate senses.
  • cluster by task.
  • assign exactly one canonical owner.
  • map supporting assets.
  • flag missing, duplicate, and uncertain clusters.
  • approve priority cluster portfolio.
Phase 5 Core information products
  • Select highest-value decision and commercial clusters.
  • create page specifications.
  • choose task-native formats.
  • add information advantage.
  • create evidence.
  • connect next actions.
  • implement technical requirements.
  • add measurement.
  • publish in cohorts.
  • review and update rule priors.
Phase 6 Supporting graph
  • Identify prerequisite and adjacent tasks.
  • validate supporting asset distinctness.
  • create or improve hubs where useful.
  • add understanding, choice, journey, and authority edges.
  • repair orphans.
  • reduce excessive depth.
  • update links after migrations.
  • measure assisted value.
  • remove weak or redundant nodes.
  • refresh canonical registry.
Phase 7 Earned evidence and authority
  • Diagnose target cluster authority gaps.
  • create linkable reasons.
  • prioritize original data, tools, benchmarks, case evidence, and visual references.
  • segment prospects.
  • launch transparent outreach.
  • track retained relevant placements.
  • measure target visibility and brand demand.
  • recover valid lost links.
  • compound successful evidence assets.
  • recalibrate authority proxies.
Phase 8 Volumetric expansion
  • Expand approved dimensions.
  • generate candidates.
  • normalize and cluster.
  • apply hard gates.
  • sample quality.
  • score survivors.
  • stage publication.
  • monitor duplicate, index, quality, and conversion behavior.
  • pause automatically on guardrail breach.
  • scale only validated cohorts.
Exit gate
canonical model is stable
AND templates are quality controlled
AND data is accurate
AND maintenance capacity exists
AND measurement cohorts exist
AND policy gates pass
Phase 9 Structural improvements
Phase 10 Paradigm experiments
Exit gate
IF new solution class completes the task materially better
AND creates defensible value
AND has a maintenance model
AND can be measured
THEN run a bounded paradigm experiment.

Ongoing cadence

Weekly operating review
  • P0 measurement/technical blockers;
  • live experiments;
  • data-quality failures;
  • major demand or value shifts;
  • guardrail breaches.
Monthly portfolio review
  • contribution value;
  • qualified conversions;
  • cluster visibility;
  • canonical ownership;
  • technical eligibility;
  • authority and evidence;
  • refresh/retirement;
  • allocation.
Quarterly model review
  • rule performance;
  • scoring calibration;
  • audience/offer changes;
  • semantic territory;
  • structural debt;
  • paradigm opportunities.

Definition of done

change is deployed
AND change ID is recorded
AND intended technical state is verified
AND measurement is active
AND owner is assigned
AND review rule is scheduled
AND documentation is updated