Background

Advertising effectiveness measures provide tools to evaluate a brand’s advertising and give marketers an early indication of advertising performance. Correct measurement is crucial as results often determine future investment for specific executions. Advertising awareness is a common measure used to identify whether there are lasting traces of a brand’s advertising in the memory of category buyers. 

Sharp, Beal and Romaniuk (2001) put forward, as a potential scientific law, the pattern that brand users are more likely than non-users, to say they remember seeing a brand’s advertising. Since then, there has been solid empirical support for the law-like nature of the pattern (Hammer & Riebe 2006; Harrison 2013; Romaniuk & Wight 2009; Sharp et al. 2001, 2002). However, this work testing the law has mostly concentrated on two approaches to questioning:  

• Which brands of <insert category> have you seen advertising for <insert time period>? 
• Have you seen any advertising for <insert brand>?

From these questions three advertising awareness metrics can be calculated: 

• Top-of-mind recall (first brand remembered as advertising); 
• Unprompted recall (any unprompted memory of the brand advertising); and 
• Brand prompted recall (memory for the brand advertising with the brand name provided). 

What unites these three metrics is that they involve the brand name, either in the cue or the retrieved information. However, given correct branding for advertising is often reported as much lower than the optimal 100% and advertising exposure can also be remembered for specific creative executions or media channels, we wondered if the presence of the brand name in the question was creating the user bias artefact, such that the empirical generalisation disappears if the advertising awareness question does not involve the brand name. 

Therefore, we set about to specifically investigate whether the bias for brand users to remember advertising more than non-users is a real reflection of differential attention/processing of advertising, or a measurement artefact that can be avoided through using a different advertising awareness measure. 

Our approach

We examine six different advertising awareness measures, each employing a different retrieval cue: 1) top-of-mind recall; 2) unprompted recall; 3) brand prompted; 4) brand plus media prompted; 5) execution prompted; and 6) execution plus media prompted. A brief description of the cueing material for each measure is presented in Table 1.

Table 1 – Summary of awareness measures and examples of questioning techniques

The data

The analysis involved 26 different data sets, covering a wide range of conditions with 101 executions, 88 brands, 18 categories, 10 countries (Australia, China, India, Portugal, Russia, South Africa, Spain, Taiwan, Turkey and UK), with advertisements sourced from a range of different media formats (TV, print, radio, outdoor and online). The samples within each data set included category buyers only.

Results

We find across all six advertising awareness measures tested that brand users are more likely to remember advertising for their brand than are non-users. Overall, 92% of the 247 observations were in favour of brand users remembering advertising, compared to non-users, with an average difference of 18 percentage points. The average results for the six different measurement approaches is shown in Figure 1.

Figure 1 – Comparison of advertising awareness levels of brand users versus non-users

We also examined if the different types of cues included in each of the awareness measures explain the difference in scores¹. The results show that using execution-prompted cues, either visual or verbal², reduces the difference between brand user and non-user awareness levels, but prompting for the brand name or media type had no impact. This suggests that the richer cues in an execution-prompted awareness measure makes retrieval of advertising information from memory easier for both brand users and non-users, but non-users will still demonstrate lower levels of retrieval. 

Implications

Our findings suggest that the measures themselves do not enhance or mitigate this user bias. Conversely, each measure simply reflects a bias that is inherently present in buyer memory to a different degree depending on the measure’s difficulty (in terms of its difficulty as a ‘recall from memory’ task). The use of easier measures (i.e. those that provide rich memory cues) heightens response levels from both users and non-users, although a disparity in favour of users remains. In contrast, the use of more difficult measures, with sparse memory cues, will dampen both user and non-user response levels, but non-users to a greater degree.

For marketers this means that the measurement approach chosen to assess the effectiveness of a brand’s advertising should be one that is an easier memory task for all respondents to avoid exacerbating the usage bias.

We further strongly encourage marketers (regardless of the advertising awareness measure used) to always separate out responses for brand users from non-users to account for the usage bias that is ubiquitous across all advertising awareness measures. For smaller brands, this will avoid under-stating awareness levels. As the smaller group of brand users with higher awareness will be washed out at the aggregate level, where the larger proportion of non-users are likely to not remember having seen the advertising. For bigger brands, the reverse is true, with splitting responses ensuring that the impact of advertising is not over-stated due to the larger base of brand users (compared to non-users). This is also beneficial for global brands where the same brand may vary in size in different markets. Separating results will allow better comparison of awareness levels for both brand users and non-users, rather than considering aggregate results across markets, which will be affected by the variation in the size of the user base in each country.

The results of this study, combined with the vast empirical evidence that shows that for a brand to grow it must attract new customers, clearly suggests that a brand’s advertising should be developed with non-brand users in mind. This (important) group of potential customers for a brand likely has very little existing brand knowledge in memory. Therefore, any exposure to advertising has to work harder to first gain attention and build new associations with the brand in memory.  

The next stage of our research is to examine the factors that might lead to greater non-user attention, such as dual branding, to attract the attention of a second brand’s customer base, or holding back on the brand reveal to try to keep non-brand user attention for longer. Initial results say this does not matter, with users and non-users having similar advertising awareness scores in either condition (for more detail see Jenni Romaniuk’s new book on distinctive brand assets). Other research is also examining whether brand users and non-users respond differently to biometric measures. Stay tuned for results on these in the future.

¹ The overall fit of the model R² = 79%, F =105.4, p<0.001.

² Verbal ß =0.21, p<0.01; Visual ß =0.12, p<0.05.