Advertising Part 2 – The Empirics of Advertising and Sales

At long last our arduous obligation as students are temporarily over! Following up from the previous article, I will introduce some empirical findings in the literature that help us better understand ‘advertising’. I am going to split the summary of empirical findings into two articles and here we will look at advertising’s effect on a firm’s sales. The next article will look at empirical findings of advertising’s effect on market share stability.

It is important to recognize the different effects that advertising could have on sales. Firstly, is there a positive relationship between present period advertising and present and future period sales? If the answer to this question was ‘yes’ then it would support some sort of goodwill building mechanism of advertising. Secondly, we need to look at whether advertising impacts the market’s demand or if it is simply a form of competition between firms within the market.

Lambin’s (1976) empirical analysis is a very influential piece of work and we will focus on the findings of this empirical analysis. Lambin considers a panel of 107 individual brands from 16 product classes and 8 Western European countries over the 1960-70’s to elicit out advertising effects between brands and also across time. The beauty of a large panel data set is that you can control for individual, fixed effects to determine long run effects of a certain variable (in our case advertising). If a researcher did not have individual firm level data, she would not be able to reliably determine advertising’s effect on long term sales as long term sales are strongly correlated with the individual firm’s product quality. Consequently, because high product quality firms are likely to spend more on advertising, if the researcher is unable to control for individual firm specific characteristics, advertising’s supposed effect on future sales will be overstated.

The paper’s findings suggest that brand advertising has significant and positive effect on current period sales and market share. Also, by interpreting the lagged-sales coefficient as a sort of measure for goodwill (where past-sales explain current sales and imply some degree of loyalty from customers,) he reports that there is evidence for a goodwill effect of advertising.

How about the other effects of advertising? Does it increase market demand? Does it have a combative role amongst the market participants? Lambin’s study also allows him to calculate other firm responses to a firm’s change in advertising spending (i.e. reaction elasticities). Lambin indicates that advertising reaction elasticities are mostly positive over time which means that if a firm increases brand advertising, the other firms are also induced to engage in more brand advertising. Also, an interesting finding is that market share and sales is not only determined by my self-advertising but negatively related with rival’s advertising spending. However, Lambin does not offer full support of advertising increasing industry demand.

Now let’s take a look at advertising’s impact on brand loyalty. We reside a little longer on Lambin’s paper. Recall that Lambin reports evidence for goodwill. Past sales seem to have an explanatory power for present sales, implying consumer inertia. However, despite evidence of the existence of consumer inertia, Lambin fails to find a relationship between this inertia and advertising intensity. This finding suggests that the persuasive role (see previous articles for an explanation) of advertising is limited and supports a more combative role.


Lambin,J.J.(1976), Advertising, Competition and Market Conduct in Oligopoly Over Time, Amsterdam: North Holland Publishing, Co..