SOCIAL MEDIA MENTION SHARE OF VOICE CASE
Completed fictional example. Values are not measured results or a benchmark.

Brand measured: Taller Luma (fictional).
Comparison set: Taller Luma and three other fictional brands.
Defined language and market: Spanish, Chile.
Period: September 1–30, 2026.
Time zone: America/Santiago.
Sources: public posts, comments, and replies on Instagram, TikTok, and YouTube.
Unit: one brand mention per reviewed post, comment, or reply.
Inclusion rule: the item identifies a confirmed name or alias of a brand in the set.
Multiple-brand rule: when an item identifies several brands, count one mention for each.
Exclusions: posts published by the four brands, exact duplicates within the same channel, brands outside the set, and private, deleted, or inaccessible content.
Channel | Taller Luma mentions | Mentions of all four brands | Taller Luma SOV
Instagram | 80 | 200 | 80 / 200 × 100 = 40%
TikTok | 60 | 200 | 60 / 200 × 100 = 30%
YouTube | 40 | 100 | 40 / 100 × 100 = 40%
Total | 180 | 500 | 180 / 500 × 100 = 36%

Interpretation: under these rules, 180 of the 500 mentions in the corpus belong to Taller Luma.
Limits: the sample does not represent every brand or conversation; a mention is not a unique person, reach, sentiment, preference, market share, or commercial result.
Next decision: review coverage and context in each channel before deciding what to investigate; keep the brand set, period, alias dictionary, and unit consistent in future comparisons.

This worksheet’s social mention method is an editorial choice applied to a fictional corpus, not a universal standard. Iris Uitz’s 2012 historical paper, p. 10, presents the method and cites Owyang and Lovett (2010): https://ibimapublishing.com/articles/JISNVC/2012/313585/313585.pdf
AMEC places Share of Voice among communication outputs; this taxonomy does not define a single social formula: https://amecorg.com/wp-content/uploads/2024/11/Taxonomy-of-Methods-Metrics-Indicators-A3.pdf
Percentages are reference calculations, not executable formulas. Brands, the corpus, and figures are fictional; the named platforms are real.