Competitive Reality Simulation: AI Citation Is a Zero-Sum Market, Not an Absolute Score
If all domains in a query cluster improve their AI visibility scores equally, every domain's citation share stays exactly the same. AI citation probability is relative. The Competitive Reality Simulation models this — and shows you where your competitive position is actually moving.
The most dangerous misconception in AI visibility strategy is treating citation probability as an absolute score. It is not. AI systems making citation decisions are not independently evaluating each source against a fixed quality threshold — they are allocating citation slots among competing sources for a given query. Improving your entity confidence from 60 to 75 matters if your competitors are at 65. It is meaningless if they are at 80 and also improving. Relative position is the only kind of improvement that produces outcomes.
The Softmax Model of Citation Allocation
The Competitive Reality Simulation models citation allocation using a softmax function over composite signal strength across competing domains in a query cluster. The probability that domain dᵢ is cited for query cluster Q is proportional to exp(σᵢ × wQ) divided by the sum of exp(σⱼ × wQ) across all competing domains j. Here σᵢ is the domain's composite signal strength and wQ is the query-type weight vector — which signals dominate for that cluster type. This is a model, not a claim about AI system internals. It captures the essential competitive property.
Query Cluster Topology
Every domain competes in one or more query clusters classified by intent type. Informational clusters have 3–6 citation slots and low zero-sum degree — multiple sources can coexist. Commercial and comparative clusters have 2–4 and 2–3 slots respectively and very high zero-sum degree. Navigational clusters are effectively monopolistic — one exact entity match takes the slot. The simulation classifies your content into clusters, estimates the competitive density and citation budget for each, and models your position within each cluster separately.
●Without competitor audit data, the simulation uses three bracket assumptions for competitor trajectories: conservative (top-quartile competitors improve at historical benchmark rate), baseline (all above-median competitors improve at benchmark rate), and aggressive (above-median competitors improve at 1.5× benchmark rate). With competitor data from Pro-tier competitive analysis, named competitor scores replace the assumptions, narrowing uncertainty intervals significantly.
Displacement Mechanism Decomposition
Citation share changes have three distinct causes: trust decay (your own signals are deteriorating — reversible by fixing Fix Plan items), competitor improvement (the competitive set is raising its signal quality — only partially reversible by improving faster), and market saturation (the query cluster is reaching competitive equilibrium — not reversible by action, requires developing new cluster coverage). The simulation decomposes projected share change into these three mechanisms so you can see which portion of a projected decline is within your control.
Absolute vs. Relative Citation Gain
The simulation always reports both dimensions separately. Total citation frequency change equals change in market share multiplied by market volume at time T, plus current share multiplied by market volume change. A domain in a growing query cluster may gain absolute citation frequency while losing relative share — both numbers matter but they mean different things. A rising tide lifts all boats; the competitive position question is whether your boat is rising faster than competitors'. SiteNexis reports both so you never conflate market growth with competitive improvement.
See where you actually stand in the AI citation competition for your primary query clusters — not just your absolute score.
Get competitive position analysis