What is excess share of voice (eSOV)?

The calculation is subtraction: a brand’s share of voice minus its share of market. The idea has a lineage. John Philip Jones documented the relationship between a brand’s share of advertising spend and its market share in 1990, and Les Binet and Peter Field made it canonical by analyzing the IPA Effectiveness Databank, hundreds of documented campaign case studies. Their finding: brands whose share of voice exceeds their share of market tend to gain share, and brands that sustain negative eSOV tend to shrink. The much-quoted average is around half a point of market share growth per year for every ten points of eSOV.

It is an empirical regularity, not a law, and the caveats are part of the finding. The growth rate per point varies by category and by brand size; large brands extract more growth per point of eSOV than challengers do. The underlying data comes from effectiveness award entries, which skews toward campaigns that worked. And the original analyses measured share of paid media spend, in an era when voice mostly meant advertising.

That last caveat is the live one. Organic, earned, and creator content now carry a large share of a brand’s actual voice, so practitioners extend the measurement to conversation share across channels. The discipline that survives translation: measure share of voice consistently, compare it against share of market, and set budgets with the gap in view.

How this shows up in Waldo

The hard half of eSOV is a share of voice series you trust. Waldo computes conversation share per brand from tracked category data, measured the same way every day, so the SOV side of the equation stays consistent; pair it with your market share data and the eSOV trendline follows.

Questions teams ask

Related terms and reading