The Transmission of Advertising by Word of Mouth

Abstract

Part of the sales effect of advertising arises through its transmission into word of mouth (WOM). Evidence on the scale of this transmission is considered and an argument is made that barriers to the carryover of ad effect into WOM may operate when a category lacks a precise form and when an ad lacks content that can be used directly in WOM. The transmission of the ad message into WOM should be an objective of those creating ad copy and this transmission should be measured in ad pretests.

Introduction

The sales effect of advertising, when there is one, can be divided into a direct effect on receivers and an indirect effect when receivers influence others through word of mouth (WOM) or when others copy the receivers’ ad-induced consumption (observational learning). These indirect effects are sometimes substantial, which should concern those who design and pretest ads.

The effect of observational learning is clearly important where categories are easily seen, such as fashion and automobiles, but this review is restricted to the way in which ads induce WOM. Advertising is a recognized trigger of WOM (Bayus, 1985; Givon & Horsky, 1990; Graham & Havlena, 2007). Lovett et al., (2019) found only small effects but Keller and Fay (2012a) reported that about 25% of WOM is stimulated in this way and Hogan et al., (2004) illustrated how measured ad effect could be much increased when it included WOM carryover. Below, findings from two studies are presented which suggest that WOM carryover from ads may be small in the case of services but quite large in respect of some durable categories (East et al., 2015; Lomax & East, 2016). To increase the effectiveness of advertising, we need to understand how WOM carryover comes about so that ads can be designed to produce it and we need to pretest this aspect of an ad’s effectiveness. Keller and Fay (2012b) have suggested that a key goal of advertising should be to foster brand conversations but no studies have been found that relate advertising content to WOM. Ad pretests that measure WOM potential are also missing.

The importance of the WOM carryover from advertising depends on how many sales are initiated by WOM. A 2005 survey (mainly of services) estimated that WOM was the main source of information for 31% of first-purchase decisions and advertising was responsible for 14% (East et al., 2005). One recent estimate of this carryover was an average of 13% (Marketing Tech Alliance, 2022). Another report stated that 20%–50% of purchasing decisions were driven by WOM (Webfx, 2022). However, these disparate averages obscure the differences between categories and category groupings. It is likely that service decisions depend more on WOM than goods decisions because advice serves as a substitute for a personal test which is often not feasible in the case of services whereas goods can often be evaluated by observation and trial. There may also be decisions, for example on hotels and holiday destinations, that are made mainly through electronic WOM (eWOM), such as internet reviews, because these are so accessible. This evidence shows that WOM will be a major basis for consumer decisions in many categories and, because of this, the role of advertising in stimulating WOM is important.

Evidence

First, data from the two studies previously cited are presented showing the proportion of WOM stimulated by advertising for eight service and four durable categories. Table 1 shows this evidence which was obtained using surveys employing a tenfold classification of triggers established by Mangold et al. (1999). The Table shows only the percentages attributed by the receivers of WOM to the advertising received by those who gave them the WOM. (Other triggers were communicator’s satisfaction, third party need, coincidental communication, receiver’s felt need, observation of other’s decision making, two or more persons deciding, receiver’s satisfaction, advertising/promotion about another product, and ‘other’).

Table 1.

Percentage of Advertisements Triggered by Advertising for the Brand.

From East et al., (2015)		From Lomax and East (2016)
Service	Percentage triggered by ads for services	Durable	Percentage triggered by ads for durables
Restaurant	4	Car	47
Mobile airtime	5	Vacuum cleaner	9
Financial services	6	Mobile phone	75
Holiday destination	3	Computer	31
Fashion shop	3
Coffee shop	3
Film	1
Hotel	4
Mean	4		42

Table 1 shows that ads for services rarely trigger WOM. Ads for the four durables trigger more WOM than services in all cases but the percentage varies substantially. This evidence shows that there are categories where carryover into WOM is likely to be responsible for a substantial part of the ad effect but that this transmission varies widely across categories. Lomax and East suggest that the physical form of durables may assist people to understand the function of these goods and to describe their benefits compared with services. Services have a variable form and the benefits derived from them may be more uncertain. Thus, there may be a transmission barrier to WOM production in the case of services. Lomax and East also point out that, if services are less supported by advertising than durables, other triggers will tend to crowd out the effect of advertising in producing WOM.

This focuses attention on the way in which the ads carry over into WOM. In a paper by East et al. (2017), which is directed mainly at the manner in which those receiving WOM may then pass on the message to others as WOM, the authors point out that this is easily done because received WOM provides a script (Schank & Abelson, 1977). This script may be used verbatim because it is already in the form required for WOM. East et al. suggest that, compared to WOM received, ads may not have the detail or form needed for an effective WOM message and the receiver may have to rework the ad content to create WOM. This would be another transmission barrier. East et al. suggest that this is likely when the WOM has to be derived from visual ads, which may lack the verbal descriptions that could be used in WOM. Another transmission barrier may occur when an ad mainly expresses emotional support for a brand and lacks appropriate descriptive terms. In consequence, there may be ads that have a strong effect on the receiver, but which lack the content to carry over into WOM. Pretests need to take this into account. Against this argument, it may be suggested that ads may sometimes stimulate receivers to restate WOM that they have previously given, so no new script is needed. Clearly, testing is required to clarify these matters.

Research Needed

Thus, we need answers to the questions:

• Are there transmission barriers? Why is there little WOM carryover from ads about services compared to durables? Is it because services receive less advertising support, because the ads are not easily reformulated into advice (transmission barrier), or is some other factor involved? It is often claimed that many services present more risk when their form is variable and thus less certain and this could affect the content of WOM about services. One consideration is the frequency of purchase with recent purchase likely both to enhance the attention to ads and increase WOM; this might affect service and durable purchases differently. Another factor is that some advertising may be designed to confirm a purchase decision already made and, as such, is likely to have content suitable for inclusion in WOM. Again, this could differ between services and durables. These questions should be answered with respect to specific categories and category groupings. For example, we need to know how ads for experience services, such as television programmes, differ in their WOM carryover from other services. The data on durables indicated substantial variation in this grouping and we need some explanation for this. Answers to these questions relate to advertising strategy; for example, can ads for services be formulated so that they lead more easily to WOM or is this likely to be a misplaced effort?

• How is WOM carryover and its effect on sales measured? Ads have been tested on the receiving individuals using brand recall, ad recall, persuasion, ad liking, eye-tracking and simulated shop purchase but these methods do not measure WOM transmission. Ads are also tested by Information Resources Inc, using the BehaviorScan split cable method (Keller & Fay, 2012a); this measures the purchase in community stores by those who have received the TV ad in their home (Hu et al., 2007). Here, it would be interesting to see whether those who have not received the ad also increase their consumption and by how much. This could indicate WOM/observational learning transmission but it may also reflect the effect of extra marketing (display, distribution) delivered to support the advertising. One approach is to measure daily WOM when there has, and has not, been an ad transmission using the method developed by Keller and Fay. The difference would give the WOM stimulated by advertising. However, this would show the relative volume of WOM when there is, and is not, advertising. Attention must also be given to differences in WOM content, and this might be done with a questionnaire formulated from qualitative research.

• What will increase the WOM transmission of ads? This paper has suggested that there may be a lack of content suitable for creating WOM in some ads. Another concern is the extent to which this content should be factual rather than emotional in form. WOM is often factual, for example, ‘X is cheaper than Y’, ‘You can get it at Z store’, ‘You can do 500 km before recharging’. It is not clear whether this factual form is a dominant aspect of WOM. But, even if it is the common form, facts will not be needed in the ad when it stimulates previously made WOM. Also, for some products, expressions of sentiment in an ad may lead to correspondingly emotionally based WOM, for example, sentimental trailers to movies could lead to WOM expressing liking of the product to others.

In Conclusion

This brief review has identified a problem in the design and testing of ads which market researchers can address by conducting research on the extent to which WOM content is derived from advertising and by developing measures of this carryover.