Salience, in marketing, is the propensity of a brand to be noticed or to come to mind in buying situations. The definition belongs to Jenni Romaniuk and Byron Sharp of the Ehrenberg-Bass Institute at the University of South Australia, who set it out in the journal Marketing Theory in 2004. The idea exists because knowing a brand is not the same as thinking of it at the moment of choice. A drinker may recognise dozens of beer labels yet, at a bar after a football match, retrieve only two. The rest are effectively absent.

Psychologists use the word for anything that stands out, and Google uses it for how central an entity is to a document; both are covered further down. The dominant usage in advertising planning is brand salience, and it is often used interchangeably with mental availability, the term Sharp popularised in How Brands Grow in 2010.

How salience works in memory

The model rests on associative memory. A brand is a node in a network of links, and each link connects it to something a buyer might be thinking about: a need, an occasion, a place, a time of day, a companion. Ehrenberg-Bass researchers call these triggers category entry points (CEPs). For soft drinks, a CEP might be "thirsty after exercise", "a family gathering" or "something to go with a takeaway". A brand linked to many CEPs, in the minds of many category buyers, has more chances to be retrieved than one linked to few.

Salience is therefore cue-dependent and competitive. Romaniuk and Sharp argued in 2004 that the traditional proxy, top-of-mind awareness, uses only one cue, the category name itself, and so misses most of the situations in which purchases happen. They proposed sampling a representative set of cues, recording yes or no links between each cue and each competing brand, and reading salience as a brand's share of those links.

Two related ideas complete the model. Physical availability is the ease of finding and buying a brand, from shelf presence to delivery coverage. Distinctive brand assets - colours, logos, characters, sounds, slogans - do not create salience on their own, but they make sure that when advertising is seen, it refreshes the right brand's memory structures rather than a rival's. Romaniuk set out the method for building and testing them in Building Distinctive Brand Assets, published by Oxford University Press in 2018.

How it is measured

The Ehrenberg-Bass approach, described in Romaniuk's Better Brand Health (Oxford University Press Australia, January 26, 2023), produces three headline numbers from a CEP survey. Mental market share is a brand's portion of all brand-CEP links in the category. Mental penetration is the percentage of category buyers who link the brand to at least one CEP. Network size is the average number of CEPs a brand is linked to among people who link it to any. Sharp has proposed a validity check: the distribution of survey responses should resemble the NBD-Dirichlet pattern found in repeat-buying data.

Most brand trackers use cheaper substitutes. Aided and unaided awareness questions remain the norm, and brand lift studies measure short-term changes in awareness, ad recall and consideration between exposed and control groups. Kantar's BrandZ framework scores brands on whether they are meaningful, different and salient; Lindsay Gorton-Lee of Kantar wrote on June 9, 2023 that "meaningful and salient" explains 80% of what makes someone want to buy a brand.

Behavioural proxies are gaining ground. Les Binet's share of search, a brand's portion of branded search volume in its category, is often read as a salience indicator, and an Institute of Practitioners in Advertising (IPA) think tank reported in October 2021 that it tracked share of market closely across 30 cases. Google now reports attributed branded searches after YouTube campaigns and claims an average 31 dollars in sales for each additional branded search, a figure Google has not published a methodology for.

Origin and evolution

The concept was borrowed from psychology. Shelley Taylor and Susan Fiske's 1978 chapter on "top of the head phenomena" showed that whatever is perceptually salient tends to be judged more causally important. Joseph Alba and Amitava Chattopadhyay took the idea into consumer research with "Salience Effects in Brand Recall", published in the Journal of Marketing Research in 1986, showing that cueing some brands could suppress recall of others.

Kevin Lane Keller put salience at the base of his customer-based brand equity pyramid. In a 2001 Marketing Science Institute working paper (report 01-107), he wrote that brand salience "relates to aspects of customer awareness of the brand", and split it into depth, how easily a brand is recalled, and breadth, "the range of purchase and consumption situations in which the brand comes to mind". The breadth dimension anticipated CEPs.

Romaniuk and Sharp's 2004 paper broke salience away from awareness and attitude altogether, arguing that awareness "should not be a key objective" for established brands. Sharp's How Brands Grow (2010) then made mental and physical availability the engine of penetration-led growth, an argument PPC Land has examined in its explainers on purchase frequency and aisle penetration. Ehrenberg-Bass work in 2012 also recast much apparent defection as light buyers who do not purchase every year.

Les Binet and Peter Field supplied complementary evidence from IPA Effectiveness Awards data. Their 2013 study, The Long and the Short of It, found that "fame" campaigns delivered 1.4 points of annualised market share efficiency per 10 points of excess share of voice, against 0.3 for others, and recommended roughly 60% of budget for brand building.

Why it matters for marketers

Salience links media budgets to long-term revenue and pricing. A Google report written with Nielsen, Kantar and WARC, The Effectiveness Equation, associated a 1% rise in brand awareness with a 0.6% long-term sales lift, against 0.4% short term, and drew on Kantar BrandZ data from 52 UK and 35 German categories that listed salience among the drivers of pricing power.

The cost of building it is rising. Charlie Ebdy of Omnicom Media Group UK argued in an IPA analysis that audience fragmentation makes mental availability "more expensive to secure". Business-to-business marketers have adopted the vocabulary too: LinkedIn's B2B Institute described CEPs as purchase-triggering contexts where buyers activate brand memories in research published in December 2025.

Limitations and disputes

The sharpest dispute concerns differentiation. Romaniuk, Sharp and Andrew Ehrenberg reported in the Australasian Marketing Journal in 2007 that across 17 markets only about 11% of a brand's users called it "different" and 10% "unique", which they took as evidence that perceived differentiation matters less than textbooks assume. Critics argue the Ehrenberg-Bass camp has been read as dismissing positioning altogether. "Differentiation is not about uniqueness, it's about difference," Mark Ritson wrote in Marketing Week on July 31, 2023, urging marketers to pursue both. Kantar's framework sides partly with the critics: its data attributes 94% of willingness to pay to meaningful difference, not salience.

Measurement is a second weakness. CEP lists are chosen by researchers, and a poorly drawn set can make any brand look salient or weak. Survey answers reflect stated memory, not behaviour, and share of search conflates salience with news events, product problems and navigational queries. Most of the evidence comes from packaged goods. Finally, salience and sales move together, but buying a brand also strengthens memory of it, so causation runs in both directions.

Salience in machines

Search engineers use the word differently. Google's Cloud Natural Language API returns a salience score "in the [0, 1.0] range" for each entity it detects, providing "information about the importance or centrality of that entity" to the text. PPC Land's explainer on named entity recognition cites Google's example of 0.79 for one entity and 0.03 for another in the same article. As of October 2026, the v2 Entity reference lists no salience field; the score survives in v1. Search engine optimisation practitioners borrow the score to judge what a page is "about", though Google has not said its ranking systems use it.

AI assistants have pushed the two meanings together. Kantar's 2026 trends report warned that "salience alone won't make brands algorithmically preferred", and that a brand absent from a model's knowledge is unlikely to be recommended by AI agents. WPP Media argued in June 2026 that the salience deciding whether a chatbot names a brand will still be built in streaming television, live sports and premium publishing.

Not the same as

Brand awareness asks whether people know a brand when prompted; salience asks whether they think of it unprompted, in a specific buying situation. Recall, and its top-of-mind variant, is one cue-specific measure of salience, usually cued by the category name. Share of voice measures a brand's portion of category advertising; the share of voice explainer describes how it is read against market share. It is an input that may build salience, not salience itself. Prominence is one of the three factors in Google's local ranking documentation, describing how well known a business is online and offline. In vision science, visual saliency, formalised in Laurent Itti, Christof Koch and Ernst Niebur's 1998 model, describes what draws the eye, closer to what attention metrics vendors measure.

Recent developments

AI search has become the main testing ground. An analysis of ChatGPT network traffic in July 2026 found that brands named in ChatGPT's own search query reached the final answer 68.9% of the time, against 2.1% for brands retrieved but never named, based on a single account. An Ahrefs experiment with a fictitious brand, Xarumei, showed AI systems repeating fabricated third-party claims. Whether brand salience among people translates into mentions by models is, as of October 2026, an open empirical question.

Timeline

  • 1978: Shelley Taylor and Susan Fiske publish their chapter on salience and "top of the head phenomena" in social psychology
  • 1986: Joseph Alba and Amitava Chattopadhyay publish "Salience Effects in Brand Recall" in the Journal of Marketing Research
  • November 1998: Laurent Itti, Christof Koch and Ernst Niebur publish their saliency-based model of visual attention
  • 2001: Kevin Lane Keller places brand salience at the base of the customer-based brand equity pyramid in Marketing Science Institute report 01-107
  • 2004: Jenni Romaniuk and Byron Sharp publish "Conceptualizing and measuring brand salience" in Marketing Theory
  • 2007: Romaniuk, Sharp and Andrew Ehrenberg report that about 11% of users call their brand "different" across 17 markets
  • 2010: Byron Sharp publishes How Brands Grow with Oxford University Press, popularising mental and physical availability
  • 2013: Les Binet and Peter Field publish The Long and the Short of It, finding fame campaigns most efficient
  • 2018: Romaniuk publishes Building Distinctive Brand Assets
  • October 2021: IPA think tank reports share of search tracking share of market across 30 cases
  • January 26, 2023: Romaniuk publishes Better Brand Health, formalising mental market share, mental penetration and network size
  • June 9, 2023: Kantar attributes 80% of desire to buy to meaningful and salient brand perceptions
  • July 31, 2023: Mark Ritson argues in Marketing Week for pursuing both differentiation and distinctiveness
  • September 26, 2024: Google developer forum users report the salience score missing from Natural Language API v2
  • March 2025: Google's The Effectiveness Equation links a 1% awareness rise to 0.6% long-term sales lift
  • November 2025: Kantar warns that salience alone will not make brands algorithmically preferred
  • December 2, 2025: LinkedIn's B2B Institute publishes research on category entry points in B2B search
  • January 8, 2026: IPA analysis by Charlie Ebdy says fragmentation makes mental availability more expensive
  • June 16, 2026: WPP Media argues salience for AI recommendations will be built in high-engagement media
  • June 29, 2026: Google introduces attributed branded searches for YouTube brand campaigns
  • August 10, 2026: Suganthan Mohanadasan publishes his analysis of brand mentions in ChatGPT search queries

Summary

Who: Jenni Romaniuk and Byron Sharp of the Ehrenberg-Bass Institute defined brand salience as a measurable construct; Kevin Lane Keller, Les Binet, Peter Field and Kantar shaped how agencies and brands use it, while Google applies the word to entity scoring in its Natural Language API.

What: The propensity of a brand to be noticed or come to mind in buying situations, measured through links between the brand and category entry points, and distinct from prompted awareness, single-cue recall and share of voice.

When: Rooted in 1970s and 1980s psychology and consumer research, placed in Keller's brand equity model in 2001, formalised by Romaniuk and Sharp in 2004, popularised by How Brands Grow in 2010, and given standard metrics in Better Brand Health in January 2023.

Where: In brand tracking surveys, media planning and effectiveness research across consumer and business-to-business categories, and increasingly in debates over how brands surface in search engines, chatbots and AI shopping agents.

Why: Buyers rarely deliberate between every brand they know; they choose among those that come to mind. Salience explains why reach-based advertising can drive growth and pricing power, though disputes over differentiation, survey design and causation limit how precisely it can be measured.