01 F1–F5 framework
Visibility is engineered, not guessed.
Answer engines don't reward effort, they reward structure. The method below is the order in which structure has to arrive — and each step has an output you can check.
- Five
- Four
- Six
- F1 audit
02 The sequence
One graph, moved through five states.
These are not five services that happen to be numbered. They are five conditions your brand's knowledge graph passes through, in this order, because each one is a precondition for the next.
- F1 AI Visibility Audit Understand how AI systems currently see your brand. Phase F1
- F2 Entity & Knowledge Hub Build the entity architecture AI systems need. Phase F2
- F3 Content Optimization Engineer content for machine comprehension and citation. Phase F3
- F4 Authority Building Establish the authority signals that drive recommendation. Phase F4
- F5 AI Visibility Monitoring Track and maintain your Share of Recommendation. Phase F5
03 Phase detail
What each phase takes in, and what it hands over.
- F1
AI Visibility Audit
Where do we appear today, and where do we not?
We measure how the answer engines describe your brand right now, across a repeatable prompt set rather than a handful of spot checks.
30–40 prompts × 4 models × 3 iterations
- Brand, category and competitor set
- Target questions and buying intents
F1 in full- Share of Recommendation, Citation Share, Top3 and co-mention baselines
- Entity disambiguation map
- Prioritized gap list
- F2
Entity & Knowledge Hub
Does the machine know who you are and what you sell?
We build the verifiable claims and the entity architecture an engine needs before it will describe you accurately, let alone recommend you.
One canonical URL per concept
- Product and service facts
- Proof, credentials and sources
F2 in full- Canonical entity definitions and relationships
- Structured data and knowledge hub pages
- One definition per concept, reused everywhere
- F3
Content Optimization
Can a machine lift an answer straight out of the page?
We make content extractable: a question, an answer that stands alone in two sentences, then the context — plus the schema, tables and comparisons engines actually quote.
Question → answer in two sentences → context
- Existing pages and content gaps
- Question sets from F1
F3 in full- Rewritten pages in the extraction pattern
- Schema, comparisons, listicles and FAQs
- Internal links that state the relationship
- F4
Authority Building
Does anyone else confirm it?
External signals and digital PR, written so third-party sources repeat the same three facts about you that your own site states.
Name + category + differentiator, every time
- Positioning and differentiator
- Target publications and directories
F4 in full- Consistent third-party mentions
- Corroborating sources engines already trust
- Improved co-mention set
- F5
AI Visibility Monitoring
Is it holding, and against whom?
Monthly tracking and a quarterly competitive comparison, using the same prompt set as the audit so every number stays comparable.
Monthly tracking, quarterly comparison
- The F1 prompt set
- Competitor watch list
F5 in full- Monthly movement in the four metrics
- Quarterly competitive comparison
- Next actions, ranked
04 Why this order
Three reasons the sequence is not negotiable.
- 01 Sequential
- Each phase depends on the one before it. Optimizing content before the entities are defined just spreads the ambiguity further.
- 02 Measurable
- Every phase reports into the same four numbers, gathered the same way, so progress is comparable month to month.
- 03 Compounding
- Structure makes content extractable; extractable content earns citations; citations feed authority. The later phases pay back the earlier ones.
05 Answers
About the method.
Do we have to run all five phases?
No. Every engagement starts at F1 because the audit sets the baseline, but where you go next depends on what the audit finds. Some brands need entity work first, others need content or authority.
How long does a phase take?
The free audit comes back in five to seven working days. A full F1 engagement — competitor set, prompt design, the whole measurement — takes two to three weeks. The build phases after it run in cycles rather than as one-off projects, because answer engines re-crawl and re-evaluate continuously.
Why is the prompt set fixed?
Because comparability is the whole point. Running the same 30 to 40 prompts across four models and three iterations means a change in the numbers reflects a change in your visibility, not a change in the test.
What if our category has no clear competitors in AI answers?
That is a finding, not a problem. Where an engine has no confident recommendation set, the cost of becoming the default answer is much lower — and F2 usually does most of the work.
06 Start at F1
Every engagement begins with the baseline.
The free AI Visibility Audit runs the F1 measurement on your brand and reports what the six engines currently say.
- Platform-by-platform visibility snapshot
- Entity clarity assessment
- Competitor comparison
- Prioritized actions