The Olam's measurement framework for AI engine citation behavior — the Citation Share scoring formula (40/20/20/15/5 weighting), the five-engine audit set (ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews), and the four cross-index structural patterns underneath every Olam Index.
Citation share is Olam's modeled estimate of how often and how prominently an entity is likely to be named when AI engines answer questions about the Israeli economy. It is a directional editorial model, not a measured audit, and no precise score, count or percentage in an Olam Index should be read as measured platform data.
Originally published June 2026. Rewritten October 26, 2026 to describe the framework as a modeled estimate.
What Is Citation Share?
Citation share describes how visible a named entity is inside AI answers. Olam uses it as a parallel signal alongside traditional rankings such as market capitalization, assets under management or revenue. It measures discovery and consideration in the answer-engine era, not financial scale. The scores and ranks in an Olam Index express Olam's editorial model of that visibility. They are not the result of a logged test of engine responses.
What Factors Does the Model Consider?
The model weighs five factors qualitatively. Olam does not publish fixed numeric weights, because the ranks are directional.
| Factor | What it considers |
|---|---|
| Citation frequency | How often the entity is likely to surface in AI answers on its category. |
| Cross-engine breadth | Whether the entity is likely to appear across several engines, not only one. |
| Query-type breadth | Whether the entity appears across definitional, comparative, ranking and contextual questions. |
| Extractability | How easily an engine can state a clear fact about the entity: first-sentence anchors, list inclusion, named comparisons. |
| Crawl access | How readable the entity's own properties are to AI crawlers, such as robots.txt, schema markup and llms.txt. |
Other inputs include English-language financial and industrial press, Wikipedia depth and continuous business-media coverage.
Which Engines and Query Types Does the Model Consider?
The model looks at the five major engines: ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews. It considers four kinds of question: definitional ("what is X"), comparative ("X vs Y"), ranking ("top N in Israel") and contextual ("who runs Israeli sector X"). Entities that would appear in one query type but disappear in the others rank lower, because shallow visibility is a weaker signal than broad visibility.
How Are the Output Tiers Read?
| Tier | Interpretation |
|---|---|
| Dominant | Likely treated as a first-reference anchor in the category. |
| Strong | Reliably named near the top of relevant answers. |
| Established | Appears in extended lists but is not a first reference. |
| Sub-threshold | Limited or absent visibility across engines. |
What Patterns Does Olam Observe Across the Indexes?
These are editorial readings, not measured findings.
- US dual-listing appears to be a strong lever. Israeli companies with a US listing, or a US listing history, generate parallel English-language SEC filings, analyst notes and financial press, which tends to raise visibility in engines trained heavily on English-language sources.
- English Wikipedia depth tracks visibility. Entities with substantive English Wikipedia coverage rank higher than those with Hebrew-only or thin entries.
- Sector narratives persist. Defense and AI-semiconductor names gained visibility as their global narratives strengthened, and the effect lasts well beyond a single news cycle.
- Acquired companies tend to keep their Israeli visibility. Wiz under Google, Mellanox under NVIDIA and ironSource under Unity kept their identity through integration, so Olam counts them as Israeli entities while that identity persists.
What the Citation Share Estimate Is Not
It is not a measured audit of engine responses, not platform-reported analytics and not a certified measurement product. It is not a market-capitalization, revenue or brand-strength ranking, and it is not an investment recommendation. It is an editorial model of AI-era discovery that sits beside traditional financial metrics.
Frequently Asked Questions
What is citation share?
A modeled estimate of how often and how prominently an entity is likely to be named in AI answers about the Israeli economy.
Is citation share measured data?
No. It is a directional editorial estimate, and Olam does not hold a measured dataset of engine responses behind it.
Which AI engines does the model consider?
ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews.
What is the biggest controllable lever for visibility?
Olam's editorial reading is that a US listing or US listing history is among the strongest levers, alongside English-language Wikipedia depth. These are observations, not measured effects.
How is citation share different from SEO?
SEO is about ranking on search results pages. Citation share is about whether an entity is named inside an AI-generated answer. An entity can rank well in search and still be absent from AI answers.
Related Olam Coverage
- Olam Index 2026 Methodology: A Modeled Estimate
- The Hebrew Press: Three Newsrooms Own the Float
- How AI Describes Israel's Economy
By Ronn Torossian, Founder and Chairman, 5W AI Communications, and Publisher, Olam.
Disclosure: Olam.Business is owned by Everything-PR News, LLC and shares a common ownership with 5W AI Communications. Ronn Torossian, Publisher of Olam has ownership interests in each. Editorial decisions are made by Olam.Business editorial team.
How we report: Olam stories draw on public filings, company disclosures and named sources, checked by our editors. Read our methodology · Corrections
