---
id: "question-som-volatility"
type: "open-question"
source_timestamps: ["§ The Rise of “Share of Model”"]
source_url: "https://hbr.org/2025/06/forget-what-you-know-about-seo-heres-how-to-optimize-your-brand-for-llms"
source_title: "Forget What You Know About Search. Optimize Your Brand for LLMs."
tags: ["metrics", "longitudinal-tracking"]
related: ["concept-share-of-model", "action-measure-som"]
resolution_path: "Longitudinal tracking of SOM metrics across major LLM version releases to measure stability and volatility."
sources: ["geo"]
sourceVaultSlug: "hbr-seg-geo"
originDay: 3
articleStem: "hbr-ext-10-optimize-brand-for-llms"
sourceUrl: "https://hbr.org/2025/06/forget-what-you-know-about-seo-heres-how-to-optimize-your-brand-for-llms"
sourceTitle: "Forget What You Know About Search. Optimize Your Brand for LLMs."
---
# How volatile is Share of Model (SOM) across model updates?

**Open question:** The article provides point-in-time snapshots of [[concept-share-of-model-d10|SOM]] (e.g., Ariel's 24% on Llama, [[entity-chanteclair|Chanteclair]]'s 19% on Perplexity), but does **not address how frequently or drastically these numbers shift** when an AI company releases a new model-weight update (e.g., GPT-4 → GPT-4o).

**Resolution path:** Longitudinal tracking of SOM metrics across major LLM version releases to measure stability and volatility.

**Enrichment:** External guidance (Agile Brand Guide, Symphonic Digital) urges brands to treat SOM figures as **directional trends, not precise metrics**, and to **re-measure repeatedly across versions** (GPT-4 vs GPT-4o, Gemini Pro vs Flash, Claude 3 vs 3.5) — each update can significantly alter SOM. Longitudinal dashboards (e.g., Shareofmodel.ai) exist specifically to manage this volatility. This reinforces the discipline recommended in [[action-measure-som]].