Share of Voice Buys Market Share on a Delay

Buried under advertising's daily metrics is one of the few relationships in this industry that deserves the word law. Decades of cross-category evidence, from the Ehrenberg-Bass school and the IPA databank most famously, keep finding the same pattern: brands whose share of voice, their slice of category advertising, exceeds their share of market tend to grow, roughly in proportion to the excess, and brands spending below their market share tend to shrink. The relationship is lagged, unglamorous, and stubborn across categories and eras. It is also the strategic answer to a question the performance decade trained everyone to stop asking: not what did this campaign return, but how loud are we relative to the fight we are in.

The mechanism is less mysterious than the econometrics. Most buyers in any category are not buying right now; they are accumulating impressions that become preferences that become shortlists, and the brand that is present during the long not-buying phase is the one remembered when the buying moment arrives. Share of voice is a crude but honest proxy for presence in that accumulation, which is why its effects arrive on a delay measured in quarters, and why the excess-share-of-voice math is really a statement about mental availability compounding, the same slow ledger I argued for instrumenting in brand awareness is measurable, stop letting people tell you otherwise. Awareness metrics tell you the state of the account. Share of voice tells you the rate of deposits, yours against everyone else's.

The honest modern complication is measurement, because the concept was born when "category advertising" meant a countable pool of TV and print dollars. Today the pool is fragmented across auctions, feeds, creators, and retail media, and no vendor sees all of it. The workable answer is a composite, built from the windows you do have: paid search impression share and Auction Insights, which I called the auction's only honest market gauge in impression share is the only auction metric that tells you about the auction; the platforms' ad libraries and estimated-spend intelligence for social and display, wide error bars, useful trends; share of search, your brand's slice of category branded queries, which the Les Binet school has shown tracks share of voice's effects closely enough to serve as a free tracer; and the earned layer, share of answers in AI engines, which is fast becoming the strangest and most important voice metric of all, per the prompt-audit practice from what is GEO. None of these is the number. Together, trended, they are a serviceable gauge of whether you are getting louder or quieter in the rooms that matter.

Then run the law against your own position, because its strategic implications are asymmetric. For challengers, excess share of voice is the price of growth, and the empirical exchange rate, historically something like ten points of excess SOV buying roughly half a point to a point of market share a year, varying by category, is the beginning of a rational budget argument, the top-down check on the bottom-up math I built in the right marketing budget is not a percentage. For leaders, the law runs as a warning: big brands can under-spend their share for quarters while retargeting efficiency applauds, which is exactly the harvest-flattered decline I keep diagnosing in the metric going up is hiding the one going down, and the invoice for the silence arrives later, addressed to someone who will not connect it to the cause. And for everyone, the law licenses the counter-cyclical move that feels wrong and keeps working: when competitors go quiet in downturns, and the tempting cut arrives on your desk too, remember you cannot cut paid media and keep the reach:, share of voice gets cheap, and the brands that hold spend buy excess share at recession prices, a pattern the IPA cases have documented across every slump since the seventies.

A client shape to make it concrete: a challenger consumer brand at roughly 6 percent market share was, by our composite, running about 4 percent share of voice, quietly shrinking while its dashboards glowed. The board approved eighteen months at an estimated 10 to 11 percent SOV, funded partly by trimming harvest spend that the incrementality tests from you do not need a data science team to run a holdout had already convicted. Share of search moved first, up 35 percent by month seven; measured market share followed, 6 to 7.4 percent by month sixteen, inside the historical exchange rate's range. Nothing about the creative was heroic. The brand simply, finally, out-shouted its size, and the category did what the sixty years of data said it would. Share of voice is not a KPI to report. It is the terms of engagement you have chosen, whether you chose them on purpose or not.