Skip to content
SEOAIExpert

METHODOLOGY · SEOAIExpert GROWTH

A search methodology built on evidence.

We combine technical analysis, search behaviour, content, product data, authority and AI-search visibility into one connected optimisation system.

Book a Strategy CallExplore Our Approach
Stages
Six
Cadence
Continuous
Basis
Observed data
METHODOLOGY ARCHITECTUREProcess

  1. Discover

    Business, catalogue, demand.

  2. Diagnose

    Technical, content, product, AI.

  3. Prioritise

    Impact, effort, relevance.

  4. Optimise

    Ship across the system.

  5. Measure

    Visibility and AI signals.

  6. Improve

    Feed evidence back in.

System

We don't optimise in isolation.

Technical work, product data, content and authority all feed the same outcome. Treated separately, they cancel each other out.

  • A fast site with unreachable categories still fails
  • Great content on an unindexable template earns nothing
  • Complete product data without architecture stays buried
  • Authority without technical access cannot be cited
CONNECTED SYSTEMIllustrative
  • Technical SEO
  • Product SEO
  • Category SEO
  • Content
  • Entities
  • Authority

AI SEARCH VISIBILITY

Composite of technical, product, category and AI-search signals.

Process

The methodology.

Six stages, run as a loop rather than a linear project. Each stage produces something reviewable.

  1. 01

    Discover

    Understand the business, products, customers and search landscape before proposing anything.

    • Commercial priorities, margins and target categories
    • Catalogue structure, variants and platform constraints
    • Demand mapping across category, product and pre-purchase intent
    • Competitor and third-party source review
  2. 02

    Diagnose

    Identify technical, content, product and visibility gaps, with evidence for each finding.

    • Crawl, render and index coverage analysis
    • Architecture, internal links and faceted navigation handling
    • Product and category template review
    • Entity, structured data and citation coverage checks
  3. 03

    Prioritise

    Rank opportunities by impact, effort and strategic relevance so the sequence is defensible.

    • Template-level fixes ahead of page-level work
    • Unblocking existing demand ahead of creating new demand
    • Dependencies mapped so work is not repeated
    • A roadmap the client's developers can actually ship
  4. 04

    Optimise

    Implement changes across the search system rather than in one channel at a time.

    • Technical and index-control changes
    • Category architecture and internal linking
    • Product data, attributes and structured data
    • Content built for identified intent gaps
  5. 05

    Measure

    Track visibility, traffic, rankings and AI-search signals against the baseline recorded upfront.

    • Visibility segmented by template, not blended
    • Coverage measured against the demand map
    • Indexation and crawl health monitored continuously
    • AI mentions and citations tracked on a stable prompt set
  6. 06

    Improve

    Use performance data to refine the system and re-enter the loop with better information.

    • Compare outcomes against the expected impact
    • Retire assumptions the data did not support
    • Expand what worked across the rest of the catalogue
    • Re-prioritise as demand and surfaces change

Diagnostic

What we analyse.

Four analysis areas, each with its own evidence requirements. Findings are recorded with the observation that produced them.

  • TECHNICAL

    Can search systems reach the catalogue?

    • Crawlability
    • Indexation
    • Architecture
    • Internal links
  • CONTENT

    Does the site answer the demand?

    • Intent
    • Coverage
    • Depth
    • Authority
  • PRODUCTS

    Can a product compete on its own data?

    • Titles
    • Descriptions
    • Attributes
    • Structured data
  • AI SEARCH

    Can AI systems describe the brand accurately?

    • Entities
    • Mentions
    • Recommendations
    • Citations

Prioritisation

Prioritisation matters.

Most SEO plans fail on sequencing rather than ideas. Opportunities are placed by expected impact against the effort required to ship them.

HIGH IMPACT / LOW EFFORT

Shipped first — usually template and index-control fixes.

  • Canonical fixes
  • Index rules
  • Title templates
  • Internal links

HIGH IMPACT / HIGH EFFORT

Planned and sequenced with development capacity.

  • Category rebuild
  • Attribute schema
  • Migration work

LOW IMPACT / LOW EFFORT

Batched into routine maintenance.

  • Alt text
  • Minor copy
  • Redirect tidying

LOW IMPACT / HIGH EFFORT

Documented and deliberately declined.

  • Cosmetic rewrites
  • Vanity tooling
Low effortImpact increases upwardHigh effort

Loop

Measurement closes the loop.

Measurement is not a report at the end. It decides what happens next, which is why the same cycle repeats every period.

Each change is shipped with an expected outcome. When measurement contradicts that expectation, the assumption is retired rather than defended. When it confirms it, the change is extended across the rest of the catalogue.

  1. Strategy

    Where visibility should come from.

  2. Implementation

    Changes shipped across the system.

  3. Measurement

    Segmented, against the baseline.

  4. Insight

    What the data actually supports.

  5. Next priority

    Re-ranked with new evidence.

  6. Strategy, revised

    The loop restarts with better inputs.

Adaptation

Our methodology evolves as search evolves.

  • Observed, not assumed

    We describe what we can observe about how surfaces behave. We do not claim to know how proprietary models rank or select what they recommend.

  • Revisited on a schedule

    Diagnostic checks and measurement frames are reviewed as search surfaces change, and outdated checks are removed rather than accumulated.

  • Documented changes

    When the methodology changes, the reason is recorded so clients can see what evidence prompted it.

FAQ

How we work, in detail.

Six connected stages: discover the business and its search landscape, diagnose technical, content, product and visibility gaps, prioritise by impact and effort, optimise across the search system, measure visibility and AI-search signals, then improve based on what the data shows. Each stage produces an artefact the client can review.

On expected impact, implementation effort and strategic relevance. Work that unblocks pages with existing demand usually comes before work that creates new demand, and template-level fixes that repeat across thousands of URLs come before page-by-page changes.

Visibility and non-brand traffic segmented by template, keyword and category coverage against the demand map, product-level ranking distribution, indexation health, and AI-search signals such as mentions and citations tracked against a stable prompt set.

AI search is treated as an additional surface fed by the same foundations, not as a separate project. The diagnostic includes entity information, structured data and third-party references, and measurement includes whether the brand is mentioned, cited or recommended.

Answer surfaces largely depend on content that can be crawled, rendered and parsed, so technical accessibility and structured data are prerequisites rather than parallel work. Entity and authority work sits on top of that foundation, which is why both are scoped in the same roadmap.

Next step

Build a search strategy based on evidence.

We start with discovery and diagnosis, so the first roadmap you see is based on your catalogue and search data rather than a template.

Book a Strategy Call

No obligation. A senior strategist reviews your store before the call.