Background
Branding seems ubiquitous, and while marketers are accused of wanting to brand everything, advertising often fails to convey the brand. When the brand is not registered by viewers, those paid-for exposures will likely contribute little to campaign objectives or may in fact be detrimental by unintentionally promoting competitors.
Good branding execution facilitates connecting advertising exposures to the brand in consumer memory (du Plessis, 2005, Romaniuk, 2009). By branding execution, we mean the visual and audio presentation of the brand name (as part of a logo, or on pack, or spoken or sung). Branding must compete for attention, visually and audibly, throughout any ad. Video advertising research has focused more on visual branding tactics than audio ones, presumably because visual memory is superior to auditory memory (e.g., Cohen et al., 2009, Standing, 1973).
However, saying the brand name in video advertising has its benefits. Audio mentions can reach people who are not paying visual attention to ads, because even though they are not looking they may still be listening. Memory theory further suggests there are separate processors and stores in working memory for different sensory systems (Cohen et al., 1993). Therefore, audio mentions offer an additional pathway to brand memory independent of those from visual cues.
Determining how to structure audio branding to best complement visual branding will help advertisers to brand their ads more effectively. This study examines how the timing and frequency of audio mentions can affect viewer brand memory. The findings can inform creative decision-making for the betterment of brand memorability linked to advertising.
Branding in Video Advertising
Past research by-and-large supports that ‘dual mode’ branding in video ads improves brand recall (Romaniuk, 2009). That is, ads both showing and saying the brand name perform better than single mode ads. The subsequent advice is to say the brand name at least once in video ads. Beyond that, research hasn’t provided an understanding of what effective audio branding sounds like. Specifically, we would like to know how often to verbalise the brand name and when.
There is some support that more audio mentions are better (Stewart and Furse, 1986, Stewart and Koslow, 1989) but not always (Romaniuk, 2009). Some studies show three or more audio mentions improve related recall (Walker and von Gonten, 1989) for both 30s and 15s TV ads (Stanton and Burke, 1998). Considering the weight of evidence, we expect to find a positive relationship between audio branding frequency and brand memory.
The timing of these audio mentions is another consideration. Should these mentions occur earlier or later? Studies generally favour earlier branding but without separating visual and audio tactics (Stewart and Furse, 1986, Stewart and Koslow, 1989, Walker and von Gonten, 1989). Research shows that early branding strengthens the subconscious process of linking the brand to the advertising content and subsequent reactions (Baker et al., 2004). This so-called primacy effect applies to auditory memory for spoken words outside of the advertising context (e.g., Sharps et al., 1996). Hence, we expect to find a positive relationship between early audio branding and brand memory.
Going further, we sought to quantify which audio branding tactic is more important… if one had to choose between often or early. Throughout our investigation, it was also essential to control for the presence of visual branding as the more impactful mode.
Data and Analysis
This research draws on data collected by the MediaScience Lab over six-years. Results from 14 studies were combined to form a dataset of 69 ads for 51 brands in two markets (USA and Australia), spanning 14 product and service categories, including consumer goods, automobiles, and retail stores. Ad lengths varied; the most common was 30s (n=45), followed by 60s (n=14), then 15s (n=10). In total, 1,774 respondents participated across studies.
To identify visual and audio brand name execution, all 69 ads were coded independently by three trained coders on a second-by-second basis. Across ads, average visual frequency was 2 (min 1 to max 8) and average audio frequency was 1 (min 0 to max 7).
We examined two brand memory measures: brand recall and brand recognition. These measures were collected immediately after respondents had watched a program containing the ads. Brand recall was measured by free recall: ‘What brands do you recall seeing during the program today?’. Brand recognition1 was measured with a brand prompted question: ‘Can you identify brands that were advertised during your session from the list below?’. Respondents gave 9,407 brand recall responses and 2,309 brand recognition responses across ads. Across ads, average correct brand recall was 17% (min 1% to max 46%) and average correct brand recognition was 63% (min 5% to max 90%).
Both descriptive and logistic regression analyses were used to identify which audio branding tactics enhance advertising effectiveness. All analyses were conducted at the response2 level .
Results
Our first descriptive result was that all ads had at least one visual brand mention but only 59% of ads (n=41) had at least one audio mention. The 41% of ads with no audio branding provided the control against which we assessed the incremental value of adding audio branding.
Does adding audio branding improve brand recall?
Consistent with prior research, we found that ‘dual mode’ branding improved brand memory. Compared to ads with only visual branding, ads with at least one audio mention had higher brand recall (21% v 16%) and recognition (69% v 52%). Even ads with just one audio mention had higher brand recall (18% v 16%) and recognition (69% v 52%). Adding audio branding clearly provides a benefit to brand memory linked to advertising.
How often do we need to say the brand name?
Next, we removed visual branding only ads from the analysis to examine the incremental benefit of more than one audio brand mention. Ads with two or more audio mentions had higher brand recall than ads with one audio mention (24% v 18%) but similarly high brand recognition (68% v 69%).
Comparing mode and frequency directly with logistic regression, increasing audio frequency had a stronger relationship with brand recall than increasing visual frequency (β =.24, p<.01 v β =.07, p<.01), whereas it was the opposite for brand recognition (β =-.18, p=.10 v β =.17, p<.001). So, there is some benefit to increasing audio branding frequency controlling for visual branding, but this effect is limited to one type of memory measure.
When do we need to say the brand name?
Early branding is considered being present in the first third of the ad (e.g., in the first 5s of a 15s ad). Ads with early audio branding had higher brand recall than ads with late audio branding (24% v 19%) but had similarly high brand recognition (67% v 70%).
Comparing mode and timing directly with logistic regression, introducing audio branding early had no (statistically significant) effect on either brand memory measure, while early visual branding had a strong positive effect on both brand recall (β =.64, p<.001) and brand recognition (β =.36, p<.01).
We further examined brand memory scores for all combinations of branding mode and timing (Table 1). Ads with early visual branding performed better on both memory measures irrespective of the timing of audio, which is consistent with the logistic regression results. Notably, ads with early audio and early visual branding (bolded in Table 1) achieved the highest scores across memory measures. Interestingly, early audio does not make up for late visual, with the lowest recognition and recall scores of all audio and visual combinations.
Table 1: Combinations of Audio and Visual Branding Timing Tactics
Altogether the results suggest that early visual branding is most important but adding early audio to early visual branding had a small incremental benefit for both memory measures (i.e., +3pp compared to early visual paired with late audio).
Audio or visual? Early or often? Recall or recognition?
Findings specific to audio branding tactics are summarised in Table 2. Looking across the results, changes to audio branding tactics generally have a greater positive effect on brand recall than brand recognition.
Table 2: Summary of Results for Audio Branding Tactics
* Logistic regression results are statistically significant to p<0.05.
Finally, we compared all visual and audio tactics directly with logistic regression (e.g., audio frequency vs visual frequency vs early audio vs early visual). The tactics that mattered most differed across brand recall and recognition.
- Brand recall: Early visual branding was the strongest driver (β =.64, p<.001) followed by audio frequency (β =.07, p<.01). Visual frequency and early audio branding were not significant. The result indicates that primacy of visual branding and repetition of audio branding helped more viewers to freely recall the brand name from memory.
- Brand recognition: Only visual frequency was significant (β =.13, p<.01). This result shows that repetition of visual branding was the primary driver of prompted (visual) brand recognition.
The findings proved quite robust when looking across product category, purchase involvement, and ad length; largely holding across conditions.
Discussion
This study provides several practical insights on how to better execute audio branding in video ads.
The previous advice to include at least one audio brand mention is supported with our data, further growing the weight of evidence in favour of this tactic. By including just one audio brand name mention, memorability scores for ads improved significantly (practically and statistically). Having at least one audio brand name mention increases the possibility for the brand to be processed and remembered, presumably through serendipitous or increased attention for the brand at some point in the ad.
Increasing audio frequency from one to two or more mentions benefited brand recall, even when controlling for visual frequency. Only one of four ads had an audio frequency of two or more in our dataset, so there is scope for many ads to make positive changes.
Saying the brand name in the first third of ads was associated with higher brand recall but was trumped by showing the brand name early. That said, both showing and saying the brand name early had the best memory results.
Importantly, these collective findings highlight that audio branding should be treated as a complement to visual branding, not a substitute.
Branding Checklist for Advertisers
When producing a new video ad, check the following to help improve brand linkage (from most to least important considerations):
☐ Is the brand name shown early?
☐ Is the brand name said at least once?
☐ Can the brand name be said more than once? (if brand recall is important)
☐ Is the brand name shown more than once? (if brand recognition is important)
☐ Can the brand name also be said early?
We caution advertisers to absorb and apply these results thoughtfully. Saying the brand name more often and earlier improved brand recall, but only a small amount of the variance in brand recall scores was explained by these executional differences (R2<.03 across tactics, i.e., only up to 3% of variance is explained by any one tactic). Hence, the effects from changing branding tactics might be relatively minor, and other factors influence advertising outcomes, such as brand size.
Final Remarks
Our findings show that advertisers can reduce the odds of wasting money by simply saying the brand name at some point during an ad. Audio branding is an effective way to enhance viewers’ likelihood of identifying the advertised brand. Saying the brand name overcomes visual avoidance and reinforces visual branding, leading to easier-to-recall brand memories.
Direct branding (showing and saying the brand name) is the most straightforward way to get the brand noticed. That said, indirect branding cues like Distinctive Assets (e.g., symbols, characters, taglines) can also affect brand memory for advertising (Hartnett et al., 2016). This study did not consider the presence of Distinctive Assets, but future research should.
______________________________________________________________________________
1 Brand recognition was collected for only a subset of ads (n=30).
2 As opposed to conducting analyses at the ad level.
______________________________________________________________________________
