Skip to main content
The SEO + GEO Domination Engine is the most differentiated surface in the Vortex IQ AI OS. It takes a single merchant brand and runs it through twelve ordered, AI-driven steps that produce a complete, market-specific content programme: every pillar topic the brand should own, every supporting spoke article, the keyword universe behind them, the AI search-engine answer surface, and the live ranking trajectory once the content goes live. The engine lives at app.vortexiq.ai/v2/seo-engine/projects/. One brand equals one project. The project runs twelve sequenced steps. Every step produces artifacts. Every artifact is reviewed and approved before the next step unlocks - the engine never silently moves forward.
GEO (Generative Engine Optimisation) is folded into every step alongside traditional SEO. The same twelve steps that build pillar-and-spoke Google ranking also build your brand’s mention surface in ChatGPT, Perplexity, Gemini, and Claude. Run both in one operation rather than treating them as separate programmes.

​
The 12 steps at a glance

​
The eight named actions

The Vortex Agents hub surfaces eight named actions covering Steps 1 through 8. Each action is a gateway into the corresponding step and can also be triggered by natural-language phrases in Ask Viq.
1

Own Brand Content (Brand Discovery - Step 1)

The engine crawls your website, inventories your product catalogue, catalogues your existing blog, profiles your brand voice and positioning, and produces a Technical SEO baseline covering crawlability, indexability, schema coverage, mobile usability, broken-link count, and redirect chains.Typical run time is 8 to 25 minutes depending on site size. If Google Analytics 4 or Google Search Console is connected, Step 1 reads those sources to enrich the brand profile with top-traffic pages, top queries, and audience signals.Artifacts produced: brand-profile.md, product-taxonomy.json, blog-inventory.csv, tech-seo-baseline.json.
2

Competitor Analysis (Competitor Discovery - Step 2)

The engine identifies 2 to 5 real organic competitors by running SERP overlap analysis against your brand’s likely keywords, filtering out marketplaces and aggregators, and ranking the remaining sites by overlap density. It then audits each competitor’s content (pillar inventory, blog cadence, content depth, schema coverage) and keyword footprint (top-ranking keywords, branded vs non-branded mix, intent split).Artifacts produced: competitors.json, per-competitor content-inventory.csv, per-competitor keyword-footprint.csv, per-competitor strengths-weaknesses.md.
3

Gap Analysis (Content Gap Analysis - Step 4)

The engine cross-references your existing content against competitor inventories and the AI engine baseline to find three things: keyword gaps (terms where 2+ competitors rank in the top 10 but you do not rank in the top 100), blue-ocean keywords (meaningful volume, no competitor ranking strongly), and beatable pages (specific competitor URLs with thin content, missing schema, or outdated dates).Artifacts produced: gaps.json, blue-ocean.csv, beatable-pages.csv, gap-summary.md.
A keyword gap is a term where competitors rank but you do not. A blue-ocean keyword is a term where nobody ranks well. Gaps require beating the competition; blue-ocean opportunities require simply showing up with quality content.
4

Keyword Research (Step 5)

The engine submits the gap inventory to Google Keyword Planner in batches, pulls 12-month Google Trends curves for each keyword, computes Vortex IQ’s own Keyword Difficulty (KD) score from SERP signal strength, deduplicates near-variants, classifies intent (transactional, commercial, informational, navigational), and shortlists the top 1,000 keywords ranked by composite score.Artifacts produced: master-keywords.csv, top-1000.csv, per-keyword trends/{keyword}.json, keyword-clusters-preview.json.
5

Content Production (Conversational Query Mining - Step 6)

The engine mines conversational long-tail queries from People Also Ask, Reddit, Quora, forum discussions, AI-engine completion logs, and voice-search transcripts. It tags each query by buyer intent (awareness, consideration, decision) and maps it to a pillar candidate. For UK Vape World this step surfaced 25,640 conversational queries assigned across 8 pillar categories at 100% category coverage.Artifacts produced: conversational-queries.csv, query-categories.json, recommendations.md, gap-coverage.json.
6

Current Page Optimisation (Steps 7 and 10)

Step 7 - Pillar Architecture clusters every keyword and conversational query into 8 to 12 pillars, each with a hub page brief, a spoke list (typically 8 to 20 spokes per pillar), and a cluster-to-pillar mapping. Blue-ocean pillars where competitor coverage is weak are flagged for fast-track treatment.Step 10 - SEO Optimization Layer scores every URL on a composite covering Technical SEO, Content & on-page, GEO readiness, and Page experience, then produces a per-URL fix list. For UK Vape World the GEO readiness score moved from 6.7 to a target of 85 once the citation blocks from Step 9 were deployed.
7

Topic / Keywords / KD / Traffic (Position Estimation - Step 8)

The engine runs every hub and spoke through a ranking projection model and writes back a 90-day projected position alongside current position, keyword volume, keyword difficulty, confidence rating (low, medium, high), and projected monthly clicks weighted by SERP CTR curves.For UK Vape World the run projected 13,007 monthly clicks at the 90-day horizon across 3,961 keyword positions and 8 pillars. On historical projects the engine’s projected rank lands within 3 positions of actual rank for 78% of URLs at the 90-day mark when content is published as planned.
8

Keyword Position Tracker (Steps 11 and 12)

Step 11 - Deployment Calendar sequences the pillar architecture into a 13-week publish schedule - for UK Vape World that was 105 articles, 8 articles per week, with pillars ordered by projected clicks per pillar. Colour-coded calendar cards show each article’s pillar, status (Briefed, In Production, Ready, Live), estimated reach, and owner.Step 12 - Operations Pack assembles the complete runbook: per-article workflow, pre-publish checklist, post-publish playbook, monitoring KPIs, rollback policy, and quarterly review cadence - all linked to live workspace tools (Kanban, Vortex Apps, Vortex Mind, Ask Viq).

​
GEO: building for AI-search alongside Google

GEO (Generative Engine Optimisation) is not a separate module - it is integrated into every step of the engine.
Step 9 writes a GEO answer-readiness specification for every URL: the standalone-quotable paragraph format, the citation hook, and the brand mention pattern that AI assistants can quote without surrounding context. Step 10 scores each URL’s GEO readiness and flags what needs to change.
Step 3 (AI Engine Baseline) measures your brand’s current citation rate across ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude using a structured set of buyer-intent prompts. The missed-prompt list feeds directly into Step 4’s gap analysis so content is planned specifically to close the AI-citation gap.
Step 7’s pillar briefs include both the traditional SEO target (head term, schema, internal-link signals) and the GEO target: the answer-readiness format, citation pattern, and brand mention pattern. One architecture, both surfaces.
Schema markup targets are specified per URL in Step 9 - Article, Product, FAQ, HowTo, Review - with the property fields the engine expects to see populated. Correct schema is a primary signal for both Google featured snippets and AI-search citation.

​
The approval model

Every step has an approval gate. The agent runs the step and generates the artifacts. You review. If the artifacts look right, click Approve and the next step unlocks. If something is off, click Reset at the top right to re-run with adjusted inputs. The engine never moves from one step to the next without your explicit approval. Each approval is recorded with an audit note in Vortex Memory so the team can trace any content or architecture decision back to when it was made and who approved it.