Share of Search: A Leading Indicator Marketers Can Compute Themselves

What share of search measures

Share of search is a ratio. Count how many people search for your brand name over a period, count how many search for the named competitors in your category over the same period, and divide the first number by the total. The result is your share of the category's branded demand, expressed as a percentage that moves as buyers shift their attention between you and the businesses you compete with.

It is deliberately a measure of attention, not of sales. Someone searching your name has not bought anything yet; they have signaled that you are in their consideration set at the moment they went looking. Aggregated across a whole category, that signal is a running tally of who is front of mind. The premise this article examines is that the tally of who is front of mind today tells you something about who will be bought tomorrow.

The obvious relative of share of search is share of voice, the classic measure of how much of a category's advertising a brand owns. The distinction matters: share of voice is an input you buy, while share of search is an output you observe. One is what you spent to be noticed; the other is whether people actually went looking.

The leading-indicator claim, and its evidence

The specific proposition that made share of search interesting is that it does not merely correlate with market share, it appears to lead it. Les Binet examined the relationship in several categories and reported that movements in share of search preceded movements in market share, in some cases (automotive, energy, and mobile among those discussed) giving up to about a year of advance warning before the sales data caught up.

A separate industry analysis, associated with the Share of Search Council and the practitioner James Hankins, looked across a set of thirty case studies and reported that share of search accounted for around 83 percent of the variation in market share. Taken together these are the two pillars the popular version of the idea rests on: a named methodology with a lead time, and a striking figure for how much of market share it explains.

Both are real, published, and attributable. Neither is a peer-reviewed academic result replicated at scale. The 83 percent figure in particular comes from one analysis of a small number of cases, and should be read as a single study's finding, not as a settled law on the order of the empirical regularities discussed later in this piece. Share of search is promising, evidenced, and early, all at once.

Why a search signal could lead sales at all

A leading indicator is only credible if there is a mechanism behind it, and marketing science supplies a plausible one. Byron Sharp and the Ehrenberg-Bass Institute's work argues that brands grow chiefly through mental availability, the tendency to come to mind easily and often in buying situations, and physical availability, being easy to find and buy. A brand that is increasingly top of mind gets thought of first, and being thought of first is upstream of being chosen.

Search behavior is a reasonable, if imperfect, external proxy for that internal state. When a brand is becoming more mentally available, more of the category's buyers reach for its name when they open a search box, and its share of search rises. If mental availability genuinely precedes purchase, then a measure that tracks mental availability should precede sales, which is exactly the pattern the leading-indicator claim describes.

This mechanism is inferential, not proven end to end. Sharp's work establishes that mental availability drives growth; it does not by itself establish that branded search volume is a clean readout of mental availability. The link is reasonable and consistent with the observed lead time, but the article is careful to present it as the theory that would explain the finding, not as a second independent proof of it.

The company it keeps: brand-building science

Share of search sits inside a broader body of brand-building research that is worth naming, because it is what gives the metric its intellectual credibility, and because that research carries its own scale caveats.

Binet and Field's The Long and the Short of It analyzed roughly 996 IPA Effectiveness Awards case studies spanning around 700 brands, 83 sectors, and three decades, and derived the widely cited guidance that brands should weight budgets toward long-term brand building over short-term activation, at roughly a 60:40 split. Separately, the double jeopardy law, one of the most replicated regularities in marketing science, holds that smaller-share brands have both fewer buyers and slightly lower loyalty among them.

The relevance here is twofold. First, these findings are the reason a search-based attention metric is taken seriously rather than dismissed as a vanity number. Second, they share the same evidentiary limit: they were derived from large, multi-brand, multi-decade datasets, and their direct application to a single-location, owner-operated business is a reasonable adaptation, not a demonstrated fact. A metric borrowed from big-brand science inherits big-brand science's uncertainty about small firms.

How to compute it yourself

The practical draw of share of search is that, unlike a randomized experiment or a firm-level marketing mix model, it is within reach of a business with no data team. The raw material is public.

The limits, stated plainly

It is worth being explicit about what this metric is not, because the gap between the careful claim and the marketed claim is exactly where buyers get misled.

The evidence base is industry research, not peer-reviewed replication. The lead time is reported across a handful of categories, not shown to be universal. The 83 percent figure is one analysis of thirty cases. And the entire brand-building canon it draws on was built on large advertisers, leaving the behavior of share of search for a single med-spa or contractor genuinely under-evidenced. This is a case where the field needs primary data, pooled across many similar small businesses, before anyone can put a confident number on how the metric behaves at that scale.

There is also a general lesson from measurement history that applies directly. The advertising industry has repeatedly adopted a convenient, computable metric, treated it as truth for years, and only corrected course once a harder test was run: the Facebook field experiments of Gordon and colleagues found that the observational attribution methods most marketers relied on frequently disagreed with randomized ground truth, sometimes in the wrong direction entirely. A self-computable metric is attractive precisely because it is easy, and ease is not evidence. Share of search earns its place by being watched carefully, with its limits stated, not by being sold as a settled predictor.

Share of search meets share of answer

There is a live reason share of search matters more now than when it was proposed. Search itself is changing. Buyers increasingly get a synthesized answer from an AI engine rather than a list of links, which raises a parallel question. Share of search asks what share of category search volume you own. A newer question asks what share of the answers those engines generate name you at all. Call it share of answer.

Here the standard of evidence is higher still. There is no standardized, agreed methodology for measuring AI-answer visibility. Generative engines are non-deterministic, so the same question can return different answers across runs; they personalize; and they are not fully observable from outside. Any "AI visibility" figure is therefore a sample-based estimate whose credibility depends entirely on a disclosed sampling method. The founding academic work in this area, the 2024 GEO paper, is barely two years old and has no canonical methods text behind it yet.

The connective idea is that both share of search and share of answer are attempts to read demand and consideration before they show up in sales, from the outside, without privileged tracking. One is more mature than the other. A serious measurement practice holds both, labels each with its true evidence tier, and refuses to present the newer one with the false confidence of the older one.

Where it fits in the measurement stack

The right posture toward share of search is neither dismissal nor overreach. It is a cheap, self-serve, directional leading indicator with real published support and real limits, best used as one panel in a wider read rather than as a single verdict. Paired with what you can observe about your standing in classic search, the local map pack, AI answers, and reputation, it becomes an early-warning line on a dashboard, not a number to stake a strategy on alone.

The discipline that makes it trustworthy is the same discipline that makes any marketing metric trustworthy: define it precisely, compute it consistently, state its error bars, and never dress an emerging signal in the language of an established law. Measured that way, share of search is a quietly powerful thing for an owner-operated business to have, a useful early look at where attention is heading, computed from tools anyone can reach.