CLIENT REVENUE GENERATED$9,720,200+Independent consulting · Since 2008
SF AI SCORE™ / AI VISIBILITY ASSESSMENT
I look beyond whether your name appears. Is the business described correctly? Is it suggested for the right service? What sources can we check? SF AI Score is the framework I use to organise those questions.
Separate a mention from a recommendation
Check the facts and visible sources
Turn findings into priorities for improvement
Himanshu Swaraj
Direct investigation, implementation and review
18 years in search marketing
Practical judgement behind the assessment
WHAT THE FRAMEWORK DOES
SF AI Score is SEOFreelance’s conceptual framework for reviewing observable AI search visibility. It groups the assessment into six dimensions so the report can distinguish presence, evidence and accuracy.
There is no published numerical formula, weighting model or automatic calculator. The useful output is a finding you can trace back to an answer and supporting evidence.
A mention, a citation and a website visit tell us different things. I keep those observations separate, then review relevant enquiry and traffic data where available.
A single answer cannot establish how often customers see your business or whether it brings qualified leads.
Review measurement and analyticsSIX DIMENSIONS
01 / VISIBILITY
Record whether the business or relevant page appears for the questions in scope. Keep the wording and context so an absence can be investigated.
02 / MENTION PRESENCE
Capture the name and surrounding description. A brand can be listed among several options without being suggested as the right choice.
03 / RECOMMENDATION PRESENCE
Record the wording, customer need and alternatives. Check whether the service and location fit before treating the recommendation as relevant.
04 / EVIDENCE SIGNALS
Save visible citations and review relevant website pages, profiles and references. A displayed source is evidence to inspect; it does not reveal the full reason an answer was generated.
05 / PLATFORM COVERAGE
Keep findings separate for each product experience tested. Record whether search was used, along with the date and relevant context, rather than combining unlike answers.
06 / REPRESENTATION QUALITY
Check services, locations, audience and important qualifications against confirmed information. Record missing details and incorrect claims as specific issues.
EXAMPLE ASSESSMENT
Suppose a plumbing business provides emergency repairs, but an AI answer describes it as an installation-only company. The name is present; the service description is wrong.
I check the claim, find conflicting information and fix the accessible sources within the agreed scope.
Then I repeat the relevant checks and record what has changed.
I can correct website content and other information we have access to. Changes to an AI answer depend on the platform and its sources, so I report the observed result rather than promise an immediate correction.
Explore AI fact audit and fixingPLUMBING BUSINESS / WORKED EXAMPLE
Mention observedThe answer names the business.
Representation needs improvementEmergency repairs are omitted; the description says installation only.
Sources need investigationCheck the service page, business profile and any cited directory.
Action and reviewCorrect confirmed errors in accessible sources, record the updates and check again.
This example explains the assessment process; it is not a client result.
REPEATABLE CHECKS
I agree the customer questions, markets and product experiences before reviewing answers. A useful record keeps the original evidence alongside the interpretation.
I compare the description with the actual services, locations and other confirmed business information. If something conflicts, I record the claim and the sources that need attention.
Save the exact question, product experience, date, market and relevant session settings.
Keep the response, visible references, business mentions and relevant alternatives.
State what is supported, what conflicts with confirmed facts and what needs further checking.
Record completed work and repeat comparable checks. Keep changed conditions visible.
I use practical finding labels such as observed, needs investigation, needs improvement and insufficient evidence. Each label needs a reason and an evidence record.
READING THE RESULTS
Answers can differ with the question, product experience, available sources and time. I keep those conditions with the evidence and explain when a comparison is limited.
If a source is not shown, I record that gap. I do not assign an unseen source or assume I know the platform’s private decision process.
Incorrect business facts need a different response from weak service content or a technical access problem. The assessment helps choose which issue to investigate first.
Traffic, enquiries and sales are reviewed separately. Seeing the business in an answer is useful evidence, but it does not establish a commercial result.
FROM FINDING TO WORK
WORK DIRECTLY WITH HIMANSHU
My 18 years in search marketing shape how I approach the work: start with the business problem, check the evidence and make the next task clear.
Your proposal defines the questions, assessment scope, implementation, fees and review points. You can start with a focused issue or a wider visibility audit.
Meet HimanshuQUESTIONS
SF AI Score is SEOFreelance’s conceptual framework for organising observable AI search visibility into six assessment dimensions: visibility, mentions, recommendations, evidence, platform coverage and representation quality.
No. There is no automatic calculator or published numerical formula here. I organise the findings with evidence, interpretation and priorities for investigation or improvement.
No. It is SEOFreelance’s assessment framework. It is not issued by Google or another AI platform, and it does not reveal private ranking systems.
The framework provides a structure for interpreting findings. An AI visibility audit is a service with an agreed scope, evidence review and recommendations. Implementation can be included in the proposal.
DAAOMI describes the work sequence: Discover, Analyze, Audit, Optimize, Monitor and Improve. SF AI Score organises the observations used to assess visibility and business representation.
We agree the relevant experiences for your customers before starting. Each observation records the product experience, question, date and relevant conditions. Coverage depends on the agreed scope and available access.
Yes. I verify the disputed claim and investigate conflicting sources. I implement agreed changes to the website and other accessible information, then repeat the relevant checks. I cannot guarantee when a platform will change its answer.
A mention or citation does not establish a visit or enquiry. I review available analytics and lead data separately so visibility observations and business outcomes are reported clearly.
The review interval depends on the business priority, the work completed and how quickly the relevant information changes. We agree that interval in the scope rather than use one schedule for every business.
Send your website, the markets you serve and the questions or incorrect answers that concern you. I will review the scope and quote for the assessment, implementation and follow-up work required.
LET’S CHECK THE EVIDENCE
Share your website and the answers that concern you. I’ll help define what to assess and what needs fixing.
