Search + AI visibility methodology
A Practical Methodology for Search and AI Visibility
A repeatable framework for defining the market, building evidence, improving discoverability and measuring change without overstating what the data can prove.
Principles
Evidence before activity
Defined scope
Markets, audiences, engines, prompts and competitors are named before measurement begins.
Comparable baselines
Methods remain stable enough to make repeated observations meaningful.
Documented uncertainty
Generative systems vary; limitations are part of the result, not a footnote.
Actionable outputs
Findings become prioritised work with owners, dependencies and validation criteria.
The 01–07 framework
How the work progresses
Frame the market
Agree objectives, audiences, geography, products and decision contexts.
Map demand
Organise search queries and AI prompts by intent, topic and journey stage.
Measure visibility
Record rankings, mentions, citations, sources and competitor presence.
Diagnose the system
Review technical access, content, entities, extractability and authority.
Prioritise change
Rank opportunities by value, evidence, effort, risk and dependency.
Implement and validate
Ship improvements, confirm quality and monitor leading signals.
Learn and repeat
Compare like with like, record changes and update the roadmap.
Measurement architecture
Different evidence answers different questions
Search evidence
- Crawl and indexation status
- Query and landing-page visibility
- Organic traffic and conversions
- Internal-link and content coverage
AI visibility evidence
- Prompt-set coverage
- Brand mentions and recommendations
- Visible citations and source domains
- Competitor and engine differences
Prompt Matrix
A structured sampling frame
The Prompt Matrix groups realistic information needs by topic, intent, funnel stage, audience and market. It is a research instrument—not a claim to represent every possible generated answer.
Define
Document inclusion rules, language, engine access and collection dates.
Observe
Capture outputs consistently and classify mentions, citations and recommendations separately.
Compare
Use the same framework for competitors and repeated measurement.
Prioritisation
Move from findings to a roadmap
Impact
Which audience or commercial outcome could improve?
Confidence
How strong and direct is the supporting evidence?
Effort
What people, systems and approvals are required?
Dependency
What must happen first for the change to work?
Research integrity
What we document
Every formal study should state the sample, collection period, tools or interfaces used, classification rules, exclusions and known limitations.
Changes in AI products, personalisation and non-deterministic generation mean observations are time-bound.
No invented certainty
We distinguish observation from inference and avoid converting a limited prompt sample into a universal market claim.
See the Research hub →Frequently asked questions
Methodology FAQ
Is the methodology the same for every organisation?
The framework is repeatable, but scope, datasets and priorities are adapted to the market and available evidence.
How do you handle changing AI answers?
We date observations, use defined samples and compare repeated runs rather than presenting one output as permanent.
Do rankings and AI mentions use one score?
No. They are different observations and should remain interpretable before any composite reporting is considered.
What makes a recommendation high priority?
Strong evidence, meaningful impact, feasible implementation and clear dependencies.
Can the method prove causation?
Not in every case. We state when a finding is correlational, observational or an informed inference.
Where can I start?
An AI Visibility Audit or scoped SEO diagnostic can establish an initial baseline.
Methodology connections
Continue with evidence and analysis
Review developing Research and practical Insights that apply this framework.
Create a measurable baseline
Choose the next decision your evidence needs to support.
Responsible expert
Methodology by Alexander Tatyanchenko
Founder of Nalmert, SEO & AI Search Strategist. Working in SEO since 2016.
Expert profile