Method in one sentence: The Healthcare Visibility Index scores approved indicators from 0 to 5 within six equally weighted dimensions, normalizes each dimension to 16.67 points, and publishes the result with an independent evidence-confidence grade.
The methodology is designed to answer a specific executive question: can the people and systems shaping a healthcare decision find, understand, verify and act on this organization? It is not designed to reward content volume, traffic scale, media noise or the simple presence of structured data. Every indicator must connect to a defined audience, a real decision path and evidence that can be reproduced during the stated collection period.
Three principles govern the model. First, visibility is multi-dimensional. A strong search presence cannot compensate for absent clinical credibility or a broken path to action. Second, evidence confidence and performance are different variables. A score of 72 supported by narrow evidence is not equivalent to a score of 72 built on decision-grade triangulation. Third, the instrument must remain honest about boundaries. It evaluates visible and verifiable evidence, not clinical quality, patient outcomes or product efficacy.
Every assessment begins with a written scope record. The record identifies the organization type, geographies, services or products, audiences, decision paths, named competitors, AI and search environments, collection period, exclusions and the person approving the scope. Changing the scope after seeing results creates a form of score shopping. Material changes require a new version or a clearly documented restatement.
| Formula: Dimension score = (indicator points earned / indicator points available) x 16.67. Composite score = sum of the six dimension scores. If a valid indicator is not applicable, remove it from both the numerator and denominator and document why. |
No reliable evidence that the required visibility condition exists.
Isolated or materially incomplete evidence; users or systems are likely to miss or misunderstand it.
The condition exists in limited channels or for limited questions, with material gaps in consistency, proof or reach.
The condition is present and usable for the defined audience, but important competitive, corroboration or continuity gaps remain.
The condition is consistently present, accurate, corroborated and competitive across most relevant decision points.
The condition is comprehensive, independently corroborated, monitored and difficult for a competitor to displace without sustained investment.
Can priority audiences find the organization, service, expertise or solution through the channels they actually use? Evidence may include organic search presence, local discovery, professional directories, referral pathways, industry media, conference visibility and relevant third-party pages.
Can search and AI systems extract a clear, accurate answer from the organization’s content? The assessment looks for direct definitions, coherent entities, retrievable facts, explicit relationships, structured content and language that answers the real question rather than merely repeating brand claims.
Is the organization named, referenced or recommended by AI engines and authoritative external sources? Citability includes the frequency and context of mentions, the quality and diversity of sources, attribution accuracy, citation consistency and whether competitors control the sources shaping the answer.
Can material claims, expertise, evidence and leadership be independently verified? This dimension evaluates visible authorship, reviewer credentials, source quality, evidence boundaries, institutional corroboration, dated updates and the difference between claimed authority and verified authority.
Is the organization visible across the actual patient, referral, procurement, adoption or partnership journey? The question is not whether a page exists. It is whether each decision-maker encounters the proof needed at the stage where uncertainty would otherwise stop progress.
Can visibility move into an appropriate next action, such as a referral, inquiry, evaluation, appointment or partnership conversation? This dimension examines conversion paths, proof near the decision, handoffs, response systems and the breakpoints between attention and commercial movement.
A minimum of three applicable indicators should be assessed per dimension. A typical diagnostic uses five to eight. Bullzeye should publish the indicator families and scoring rules but does not need to expose client-sensitive prompts, competitor lists or proprietary collection automation. Any custom indicator must be defined before scoring and approved in the scope record.
The numeric score describes observed performance. The confidence grade describes how much trust leadership should place in the result. Bullzeye’s Healthcare Growth Intelligence standard uses three grades. Directional evidence has one credible source, limited corroboration and documented assumptions. Supported evidence uses multiple source types that point the same way, reviews contradictions and identifies material assumptions. Decision-grade evidence independently triangulates the evidence types relevant to the decision, explains material contradictions and residual risks, and carries approval from a named owner.
Confidence is not calculated by counting sources. Three weak sources do not become decision-grade because there are three of them. Relevance, independence, recency and the decision being informed matter. A public-facing baseline built from controlled tests and third-party observations may be Supported. A high-cost, hard-to-reverse repositioning decision may require additional buyer evidence, competitive evidence and commercial validation before it becomes Decision-grade.
AI and search results vary by date, location, account context, engine, model and prompt wording. The methodology therefore records enough context for a qualified reviewer to repeat the test. For each prompt or query, record the exact wording, engine, interface, date, location or market, logged-in state where relevant, response, cited sources, capture and reviewer. Repeated tests should use a stable prompt panel and a defined cadence rather than ad hoc questions selected after results are visible.
Third-party authority should also be classified. Government agencies, peer-reviewed literature, recognized professional bodies, reputable health information resources, institutional pages, earned media, directories, review platforms and self-published pages do not carry the same evidentiary value. The Index should record source type and relevance instead of treating every mention as equivalent.
| Important limitation: A score cannot be described as above average, top quartile or benchmark-leading until Bullzeye has a defined sample, transparent comparison method and adequate data. Version 1.0 score bands describe internal consistency, not market rank. |
Each methodology release should state its version, publication date, material changes and whether prior scores remain comparable. Evidence has a shelf life. AI systems change, search features evolve, physician roles change, products launch, acquisitions alter entity relationships and competitors publish new proof. A reassessment should record whether it is a comparable rescore, a scope change or a new baseline.
A dimension cannot be scored from one convenient data source. Search results can inform discoverability but cannot establish credibility. AI responses can inform answerability and citability but cannot establish that a claim is clinically sound. CRM activity can inform commercial continuity but cannot show whether an external audience could verify the organization before acting. Each dimension therefore requires a minimum evidence set appropriate to its purpose.
Version 1.0 does not prescribe an industry-wide prompt count because the correct panel depends on scope. It does prescribe how the panel is selected. Prompts must represent real audience questions across discovery, comparison, verification and action. They must be approved before the main collection run. Bullzeye should not add only favorable prompts after seeing results or remove difficult prompts because the organization performs poorly.
For an initial organizational baseline, Bullzeye should normally use at least 20 approved questions across the priority journey and test them on the engines materially relevant to the audience. That is an operating minimum for directional pattern detection, not a statistically representative market sample. Material claims about industry share, percentiles or causation require a separate research design, a documented sample and appropriate statistical review.
AI observations should be repeated because generated answers can vary. The collection record must distinguish a one-time observation from a repeated pattern. If an engine does not provide a cited source, that absence should be recorded rather than reverse-engineered into false certainty. The methodology evaluates what is observable and retains the limitation.
Contradictory evidence is not noise to be averaged away. It can reveal audience differences, unstable representation, outdated sources or a real strategic inconsistency. Reviewers should record the contradiction, identify which evidence types disagree, assess relevance and independence, and decide whether the score should be reduced, segmented or held pending further collection.
For example, a company may appear frequently in AI answers while health-system buyers report that they cannot find implementation evidence. Citability may score well while decision-path presence remains weak. A physician may have strong institutional credentials but outdated public profiles. Credibility may be supported while answerability is weak. The methodology is designed to preserve those differences.
Every completed diagnostic needs a named scoring owner and a separate approval owner. Indicators involving judgment should receive a second review. Content that makes material clinical claims requires a reviewer whose qualifications are relevant to those claims. Content about procurement, economics, security or commercialization may require a different subject-matter reviewer. A physician credential should not be used as a universal substitute for the expertise the claim actually requires.
The public page should display the real author, reviewer when applicable, publication date, modification date, source list and material limitation. JSON-LD must match what users can see. Bullzeye should add reviewedBy only after a real person or organization has approved the content and is named on-page. Schema is a description of the governance process, not a replacement for it.
The first release is a transparent operational model. Validation should proceed in stages. First, conduct internal inter-rater testing to determine whether trained reviewers score the same evidence consistently. Second, compare scores with qualitative interviews from the audiences being assessed. Third, examine whether prioritized interventions change the relevant dimension and commercial signal over time. Fourth, after a sufficient and comparable sample exists, evaluate benchmark ranges and predictive relationships.
Until those stages are complete, Bullzeye should describe the Index as a diagnostic methodology, not a validated clinical instrument, industry standard or predictive model. That limitation does not weaken the method. It prevents the brand from claiming a level of proof it has not yet earned.
Each engagement should maintain a scorebook containing the approved scope, indicator definitions, raw captures, source URLs, reviewer notes, calculations, contradictions, exclusions and approval record. The published report may summarize sensitive evidence, but Bullzeye must retain enough material to reproduce every score and explain every judgment.
Access to client-sensitive prompts, buyer interviews and commercial records should be role-based. The methodology page should explain the type of evidence collected without disclosing confidential source material. Retention periods should be agreed in the engagement scope and aligned with contractual, privacy and security requirements.
A completed scorebook should allow another qualified reviewer to trace the conclusion without relying on the original scorer’s memory. If that trace is not possible, the result remains Directional regardless of how persuasive the recommendation appears.
The public methodology page should display version number, effective date, author, named reviewers where applicable, scoring formula, indicator-selection rules, minimum evidence requirements, known limitations and change history. Any material revision to dimension definitions, weights, score bands or confidence criteria requires a new methodology version and an explanation of comparability with prior scores.
Minor editorial corrections can retain the same version when they do not change scoring or interpretation. Bullzeye should record them in the change log rather than silently replacing methodological language. This gives clients, reviewers and future benchmark users a stable basis for understanding which standard governed a result.
Equal weighting is a governance decision, not a claim that every visibility dimension has the same causal effect on revenue for every organization. The model gives each dimension the same maximum contribution so an easily measured strength cannot erase a material weakness elsewhere. A health system with strong discoverability but weak decision-path presence should see both conditions clearly. A MedTech company with strong buyer evidence but weak citability should not be penalized by an invented industry average. Equal weights keep the diagnostic stable while the scoped indicators inside each dimension reflect the organization, audience and decision being assessed.
If later research demonstrates that a different weighting model is more valid for a clearly defined population and outcome, Bullzeye can test that model as a new methodology version. It should not quietly tune weights until a desired score appears. Any proposed change requires a stated hypothesis, comparison against the equal-weight baseline, documented sample, error analysis and a change record explaining whether old and new scores remain comparable. Until that evidence exists, the six equal weights are the most transparent way to preserve the full strategic definition of visibility without pretending to have predictive precision that has not been established.
Equal weighting protects the strategic definition of healthcare visibility. It prevents the model from over-rewarding the signals that are easiest to collect, such as rankings or mentions, while underweighting credibility, decision-path presence or commercial continuity.
No. The dimensions and weights remain constant, but the indicator set is scoped to the organization type, audience and decision journey.
A legitimately non-applicable indicator is removed from both points earned and points available. The rationale must be recorded before the final score is approved.
Independent triangulation across the evidence types relevant to the decision, explained contradictions and residual risks, and approval by a named owner.
Only with caution. Direct comparison requires aligned scopes, audiences, markets, collection periods and indicator definitions. Version 1.0 is best used for baseline, gap prioritization and change over time.
No. It scores whether evidence is visible and verifiable, not whether care, outcomes or efficacy are superior.