Acknowledgment
We thank Danae Underwood for her valuable research conducted during her Master’s studies at The Ehrenberg-Bass Institute, which contributed to Study 2 in this report.
Background
Mental Availability is the propensity of a brand to be noticed and/or thought of by those in buying or consumption situations (For more background see Ch 6, 7 & 11 of Better Brand Health). Mental Availability theory builds upon the Associative Network Theories (ANT) of memory, which describes human memory as a network of nodes connected by associative links. In marketing, these links in memory are commonly referred to as brand associations. Mental Availability depends on memory, not attitude. When faced with a brand choice, whether it be ordering new running shoes, selecting a washing liquid at a supermarket shelf, or considering which restaurant to visit, buyers largely rely on their existing memories. Memory-based brand associations help consumers to “decide” which brands to see or recall at that point in time. Brands with higher Mental Availability have a greater chance of being thought of or noticed in that moment, which is the first necessary step for a brand to be chosen.
A key purpose of advertising is to maintain or grow a brand’s Mental Availability. Advertising can refresh brand linkages in memory and sometimes build new memories/links to the brand. So performance on these goals should be measured to evaluate the advertising’s effectiveness.
To measure a brand’s Mental Availability, Professor Jenni Romaniuk (2013) proposed four metrics, which can be determined from data collected across category buyers via surveys:
- Mental Market Share (MMS) is the overall quantification of Mental Availability. Providing a share of total memory associations relative to competitors.
- Mental Penetration (MPen) identifies the percentage of category buyers who have at least one association with the brand. It measures the breadth of the brand’s memory network, i.e., in how many buyers’ brains it has a presence.
- Network Size (NS) is the average number of associations a category buyer with Mental Penetration has with the brand. It measures the depth of the brand’s memory network in those who have at least one association with the brand.
- Share of Mind (SOM) is the relative share of associations for the brand, vs all others in the category, among those who have at least one association with the brand.
Category buyers are presented with relevant associations via a pick-any approach, along with metrics gauging buying behaviour at brand and category level.
Research Questions
Much Research and Development (R&D) has been devoted to developing and evaluating MA measurement against other options (e.g., Barnard & Ehrenberg 1990; Driesener and Romaniuk 2006; Romaniuk, 2006, 2008, 2023; Romaniuk and Huang 2019). While the metrics were developed drawing on generalised knowledge about memory and brand health, as well as systematic documentation of how brands compare in the associations they have and how the associations vary across conditions and time, the validity and robustness of these measures to capture short-term change in response to advertising has not previously been explored. This leads to the first research question:
- Q1: To what extent, and how, do each of the Mental Availability metrics respond to brand advertising?
Furthermore, it is known that brand users/buyers/owners possess more memory associations than brand non-users due to greater past experiences and brand exposure (Bird, Channon, and Ehrenberg 1970; Burnkrant and Unnava 1995; Krishnan 1996; Romaniuk, Bogomolova, and Dall’Olmo Riley 2012; Vaughan 2016). This is known as the brand usage bias, but how advertising exposure affects these associations in consumer memory is not well documented. Research has shown that a brands buyers/users/owners notice and recall its advertising more easily compared to those with no, or limited, knowledge of the brand (for more detail see Report 87). Given these known affects from brand usage, there is a need to research how the MA metrics capture the effect of advertising on a brand’s Mental Availability for these different groups of respondents. This leads to the following question:
- Q2: Which Mental Availability metrics are most appropriate to use to measure advertising’s effects on brand users versus non-users?
Multi-Category, Multi-Country, Multi-Method Testing
Our multi-study design with built-in replication sought to provide a comprehensive test of the MA metrics across different methods, categories and countries. We now briefly outline the data and research design of Studies 1 and 2, with references to the full studies provided at the end:
Study 1. Aggregate test: This study drew on secondary cross-sectional data which was collected for Brand Health Tracking Studies in the UK and Australia. Ten advertising campaigns, spanning eight brands from a range of categories including Financial Services, Consumer Packaged Goods (CPG), and Fast Food/Quick Service Restaurants were analysed. They utilised a variety of media (TV, radio, outdoor and online) and were on-air a maximum of four weeks. This study looked at aggregate level change over time, and while it does not directly control for ad exposure, exposure to the advertising was determined as detailed in Appendix A.
Study 2. Individual level test: This study was a forced exposure advertising experiment with a longitudinal design gathering responses from the same individuals. A sample of 3,193 category users in Australia resulted in n= ~790 respondents for each of the four categories studied: Financial Services, On-demand Streaming, Chocolate, and Coffee. Individuals were surveyed online twice with brand attribute questions: pre- and post-advertising exposure. Respondents completed Survey 2 within a window of two to seven days after Survey 1. A mid-sized target brand was selected for each category, and a real world (new to market) 30 second video advertisement shown during a road safety video (video 2min 35sec).
Key Findings
Finding 1: Mental Availability metrics detect advertising effects.
The results across both studies show mostly positive increases for all Mental Availability metrics after advertising exposure when looking at the total sample pre and post exposure (see Table 1). As expected, the gains tend to be modest, particularly in the brand tracking results from Study 1. Overall, there was positive movement in 25 of the 40 observations (10 campaigns with 4 measures) in Study 1, and 14 of 16 observations (4 ads with 4 measures) in Study 2 (which controlled for external factors more tightly). This is a robust demonstration that the MA metrics capture advertising’s effect on MA.
Table 1: Average percentage point and % difference in MA Metrics across brand level results from Wave 1 to Wave 2 [Study 1 does not control for Ad exposure]
Finding 2: The biggest change in Mental Availability is among brand non-users.
The results show that, on average, by far the greatest change across the four MA metrics is among brand non-users. In Study 1, for the MMS metric, there was an average increase of 9.4% in the share of associations for non-users, compared with only 3.8% for brand users.
However, in Study 1 we could not be sure the respondents had been exposed to advertising. In Study 2, we remedied this lack of advertising control. The change seen was a 19.2% increase in MMS % among non-users compared with only 2.3% for users.
Looking across the other three metrics for the target brands, this pattern is consistent, with the change from pre to post advertising exposure for Study 2 many times greater for non-users than users (results shown in Tables 2 and 3).
Finding 3: Share of Mind captures change in Mental Availability for brand users.
Among brand users, the SOM metric had the largest change. This metric increased, on average, by 5.1% in Study 1. The results were clearer in Study 2 where ad exposure had occurred for all respondents, with a 6.1% increase for brand users. Table 2 shows these results for each of the four categories examined.
Table 2: Study 2 Brand Users – Brand level MA Metrics in Wave 1 and Wave 2
Table 3: Study 2 Brand Non-Users – Brand level MA Metrics in Wave 1 and Wave 2
Finding 4: Mental Penetration captures the biggest difference in Mental Availability for non-users who have been exposed to advertising.
Among brand non-users, MPen saw the greatest positive change. In Study 2, all four MA metrics increased for the target brands’ non-users, on average. The MPen metric had the largest average increase of 5.4 pp after advertising exposure, and was significant for 3 of the 4 brands (see Table 3). The chocolate brand was already very high on MPen and reaching a ceiling on this metric, even for brand non-users.
We see a similar result in Study 1 when awareness of the advertising is factored in. To account for the effect of advertising in Study 1, we looked at the magnitude of difference in the MA metrics among brand users and non-users (rather than change over time), classifying respondents as being aware/not aware of the advertising campaign that had just run (i.e., recognising or not). We calculated a ratio metric to indicate how many times higher the metric was for those aware of the advertising vs not-aware.
Here, we saw MPen showing the biggest differences among non-users, with those that recognised the advertising being 1.6 times, on average, more likely to provide at least one association with the advertised brand. This was followed by MMS, which was 1.4 times greater, on average, than among those who did not recognise the advertising. We see a similar pattern for brand-users but with a lower magnitude. Full results are available in the references listed below.
Conclusions
MA metrics are an easy and useful tool which marketing professionals can use to track the effect of advertising on their brand’s Mental Availability. These two studies provide robust empirical evidence that the MA metrics are affected by advertising exposures, but the results must be interpreted with consideration of prior brand usage. As such, marketing professionals should anticipate different patterns of change in the metrics between consumer groups (users and non-users).
Brand Users
Brand users already have a high number of associations in memory for brands they use, as reflected in the Mental Availability scores. The MPen and NS metrics among brand users hit a ceiling. This aligns with Romaniuk’s (2023) research showing that MPen’s usefulness for bigger brands depends on whether there is ‘room to move’, whereby brands with Mpen over 80% are unlikely to see major changes.
It is possible that the greater movement seen in SOM, a share metric, for brand users is the result of strengthening pre-existing associations in memory, occasionally adding a new association(s) and ‘crowding out’ or interfering with the retrieval of competitor brands from memory.
Brand Non-Users
Brand non-users have fewer memories and links to the brand in question. When non-users are reached with advertising, it can help form initial memory structures for the brand (or refresh those that are already there but not fresh or as accessible). This will result in increases in MPen as well as SoM for those with some prior associations.
