Does advertising grow markets?

‘Does advertising grow markets?’ written by Bridget Angear

Sometimes advertising does grow markets, but much more often it doesn’t. #AdBigQs

It is a tendency of those of us who work in advertising to exaggerate its influence. We love to believe our own hype about its power to create consumer desire. The logic goes something like this: you advertise a product or service; as a result, more people desire it and go on to buy it (if the advertising is successful); therefore, the market in which that product or service operates gets bigger. With the ultimate conclusion being, therefore, that advertising grows markets. It’s a beguilingly simple, seductive argument.

Indeed, a poll conducted in February 2016 asked the members of the Account Planning Group (a community of planners and strategists in the UK and beyond) whether they agreed or disagreed with the statement ‘Most successful advertising grows markets’. Two thirds (67%) agreed that most successful advertising does indeed grow markets.

Advertising probably began as simple signposting. One of the oldest ‘ads’ we know of is a first-century carving on the quay at Ephesus, directing sailors to the local brothel. With printing, the ‘signposting’ became more sophisticated; it became a way of putting sellers in touch with buyers, employers in touch with jobseekers, Miss Lonely Heart in touch with Mr Right. Such paid-for notices filled the classified columns of the press from the 17th century onwards. And the Internet, of course, is an even better medium for signposting.

But wherever signposting becomes more common, and therefore more competitive, it inevitably develops into both a fight for attention and a sales pitch. Three hundred years ago, Joseph Addison was reflecting facetiously in The Tatler on the devices that advertisers in his magazine used to get their own message noticed, and the high-flown language they adopted to make their products desirable.

If this widely held opinion is correct, it has major consequences. Imagine you’re a government seeking to curtail consumption of ‘harmful’ products (such as alcohol and high fat, high sugar foods and drinks). If the argument holds true that advertising grows markets, then it should be the case that the converse also holds true. And that restricting the advertising of ‘harmful’ substances should restrict their market growth. With ramifications such as these, and temperatures running high on both sides of the ‘advertising freedom’ debate, the need for an evidence based answer to the question has never been more important.

Simon Broadbent’s seminal paper ‘Does advertising grow markets?’, first published in 1997, provides a forensic, meticulous deconstruction of the question and an empirically based answer to it. In part, Broadbent draws on the analysis carried out by Colin McDonald and Vivian Braun of the 156 case studies that feature in Volumes 1–8 of Advertising Works, the public record of the IPA’s biannual Effectiveness Awards scheme. (These volumes document ‘proven’ effectiveness cases from 1980 to 1994.)

This data set comes with a few health warnings – the case histories are mainly from the UK, with contributions mainly from large UK advertising agencies; the sample size is relatively small; the cases cannot be regarded as typical, since they are ‘by definition’ particularly successful examples of advertising; and their authors were free to choose whether or not to include market size effects. (Although, regarding the last point, one could assume that if there were market size effects, the author would have been keen to cite them.)

While Broadbent was very careful to avoid making any generalisations, I am going to be bold enough to make one.

Most advertising does not grow markets

How can I be so sure? In the vast majority (83%) of the IPA cases explored in Broadbent’s paper, there was no evidence that advertising had grown the market. In the majority of the cases, advertising was shown to have driven brand share but not market growth.

This is not to say that advertising can never grow markets. Nor that this might not be a very good thing to attempt, in the case for example of organ donation. It is just that these cases are very rare.

There were, of course, exceptions to that rule and Broadbent went on to list the six conditions most likely to produce market growth. These are worth noting should you wish to embark on a market growth strategy, since the evidence would suggest you need one or more of these conditions to succeed.

Broadbent’s conditions for market growth were as follows:

1. Small markets with very similar products
2. New markets with many products
3. A market with a high proportion of trialists
4. A market with a single dominant brand
5. A market that is already growing for other reasons, such as social change
6. Or – more straightforwardly – when a group of manufacturers combines to produce a generic campaign.

In 2008 Tim Broadbent revisited Simon’s conclusions in a paper entitled ‘Does advertising grow markets? More evidence from the United Kingdom’. This time the evidence base was 129 IPA papers from 1996–2006, which were examined by Les Binet and Peter Field of the IPA Databank. One hundred and nine of these were typical commercial cases (20 being for government, social or charity advertising, or for generic campaigns, that were excluded to be consistent with Simon Broadbent’s original investigation).

This updated data review once again supported the assertion that most advertising does not grow markets, finding this time that just 14 of the 109 cases reported market growth. Expressed another way, 87% of this case study cohort showed no evidence of market growth.

Seismic changes have ‘of course’ occurred in our industry since 2006 (most people still got their news from newspapers back then, Twitter had only just launched and adblocking was unheard of). So it seemed appropriate to once again dive into the IPA Databank to see if these conclusions still held true. Binet and Field agreed once more to review the cases, this time those written in the ten years since 2006.

The results this time around were even more definitive. Advertising was shown to have grown the market in just 13 (8%) of the 157 cases reviewed. In other words, a further decline from the 17% and 13% figures previously reported.

Looking back across our 25-year data set, then, we can see that market effects are only reported in a small minority of winning IPA cases and these effects are declining. Since these cases are the industry’s most celebrated examples of effectiveness, it seems fair to assume that the overall incidence of advertising driving market growth is even lower, and also declining.

We should point out at this stage that, while the evidence put forward in this paper supports the conclusion that most individual advertisers do not grow market size, the issue of whether the total weight of advertising in a market can grow a market has not been specifically explored and remains worthy of further investigation. But if even our industry’s most effective campaigns tend not to contribute to market growth, it is difficult to see why the aggregate effect of the individual advertising in any given market would be substantial.

We note also that generic campaigns, designed to grow market size rather than market share (for example, of meat, fish or milk) have largely disappeared from view, suggesting that their efficacy is dubious.

We can put forward a number of hypotheses as to why market effects seem to be getting smaller over time: mature markets (perhaps exacerbated by recessionary times), a lack of innovation and perhaps even marketers now being incentivised on share gain rather than market growth. But it is worth noting that none of these have yet been validated.

One question we can answer is whether the six conditions for market growth outlined in Broadbent’s original paper could be detected in the thirteen cases that did evidence some market growth in our latest data period.

It should be noted that since there were no generic campaigns among those reviewed this time around, there are no cases that support his sixth condition for market growth: ‘when a group of manufacturers combines openly to advertise for this purpose with a generic campaign’. (This is not to say that they do not work, just that there were none in this analysis.)

However, all of Broadbent’s remaining five conditions were present, to a greater or lesser extent, across the papers that did evidence some market growth. More detailed analysis suggests that two were more prevalent than the others: ‘a small market with very similar products’ and ‘a market with a single dominant brand’.

Carex, Danone Activia and Corsodyl were all examples of a small market with very similar products. Carex grew the liquid soap market by persuading people who previously had not washed their hands to do so. Danone Activia grew the yoghurts and chilled desserts category by giving women a positive ‘reason to eat’. And Corsodyl grew the medicated mouthwash market by helping to tackle gum disease.

Bisto, Camelot, Virgin Trains and the New York Bakery Co. all exemplified markets with a single dominant brand, each of which had to grow consumer penetration to realise ambitious growth targets.

It is worth noting also that product innovation played a role in most of the 13 cases of market growth – among others, for Karcher window cleaners, easyJet and Gü puddings. In fact, it could be argued that innovation is almost a prerequisite of market growth and – by extension – that innovation must be present alongside one of Broadbent’s other conditions. Indeed, to quote from Broadbent’s original paper, ‘innovation and quality improvement (in the consumer’s interest) are plausibly the more important factors [than advertising].’

Why, then, is it the case that most advertising seems not to grow market size? I can offer two possible explanations here. Firstly, that a combination of product innovation and one of the conditions outlined by Broadbent seldom occurs.

Secondly, and perhaps more commonsensically, is the simple fact that across our case studies, and in markets more generally, some kind of substitution probably occurred: that when a consumer bought more of one thing, they bought less of something similar. The result was that the overall market did not grow.

This leads me to the second lesson to be taken from Broadbent’s original paper, beneath the headline that individual advertising efforts tend not to grow markets. Namely, that the question of whether advertising grows markets cannot be truly answered until a particular market is meaningfully defined, ideally by consumers (rather, that is, than by manufacturers, retailers, research companies or even lobbyists, who will each seek to define a market to serve their own best interests), and that a careful definition of our ‘market’ has wider benefits for practitioners, as the definition chosen will affect the way advertising is planned, deployed and evaluated.

Marketers have a tendency to define markets by products of a similar nature: the car market, for example, or the soft drinks market. This is not necessarily how consumers might think of them. To the consumer, different products can meet the same need or want, and it is these needs or wants that come first when they are considering their options. As Theodore Levitt pointed out as long ago as the 1960s, people buy railroad tickets not because they want railroad tickets but because they want transportation. The alternative to a railroad ticket might not be a different railroad ticket, but a bus ticket, an airline ticket or a hire car.

Broadbent uses the following example in his paper to highlight the difficulty and subtlety of market definition. Suppose we want to know whether successful advertising for a particular car – in this example, a Mitsubishi Shogun – had grown the car market. There are many answers to the question ‘Which car market?’ and the market definition that we choose is likely to affect whether market growth can be claimed.

Common sense suggests that the further up the categories you go, the lower the likelihood of detecting market growth. And, in reality, consumers tend to define markets in much broader terms than marketers: i.e. towards the top of the ladder. The choice of whether to take a holiday may compete with a desire for a new car, or a chocolate bar with a soft drink at a train station vending machine.

Maybe then, when we consider things in this light, it is not so surprising that advertising seems to have limited ability to create new desires and, thus, fuel market growth rather than just brand growth. Its main potency lies in presenting consumers with different options to meet their existing needs and wants: something for legislators and lobbyists to consider when calling for a ban on ‘harmful’ products, as there may well be more powerful ways to encourage a desired behaviour. Reducing portion size, for example, might do more to reduce waistlines than a ban on ‘junk food’ advertising.

Using Broadbent’s terminology, the majority of our most recent IPA ‘market growth’ cases grew markets at the sub-sector level. Corsodyl grew the medicated mouthwash market, but this might have been at the expense of ‘regular’ mouthwash sales. Danone Activia grew the ‘chilled desserts and yoghurts’ category, but we do not know what happened at the total desserts level (which could of course include foods such as fruit, cheese and ice cream). The New York Bakery Co. grew the bagel category, but was this at the expense of total bread sales? Common sense suggests that at least some of their sales increases will have come from consumer substitutions at the category or supercategory level.

Depending in part on our definitions, then, we must conclude that advertising can grow markets, but doesn’t in the vast majority of cases: it typically rearranges brand share or ‘sub-sector’ share. Those of us looking to grow or shrink markets should look to other levers first.

Bridget has a degree in Economics and started her career in advertising at J. Walter Thompson. Ten years later, she left as Planning Director to join Abbott Mead Vickers BBDO where she is now joint Chief Strategy Officer. She loves nothing more than tackling big challenges. Her hallmarks are simplicity, clarity and friendliness. Bridget is also extremely proud that, in 2011, AMV BBDO won the inaugural Creative Effectiveness Grand Prix in Cannes, as well as being the first agency ever to be awarded APG (Account Planning Group) and IPA ‘Agency of the Year’ accolades in the same calendar year. She was convenor of Judges for the 2016 IPA Effectiveness Awards and a keen champion of effective creative work.