Stars in your ads: What evidence is there that celebrity sells?

Since the inception of the awards in 1980, and looking only at the main biennial competition, we can see that on average just over 14 per cent of winners used celebrity in their campaigns. However, applying a four-year rolling average, it’s clear that there’s a steadily increasing incidence of stars with the average rising to nearly 24 per cent. Marketers and their agencies are appreciating the amplifying effect a well-chosen celebrity can have in propelling their idea into the marketplace.

It’s worth noting that this sort of campaign is disproportionately successful in this the most demanding of all adland’s award schemes. For example, of the 18 IPA Grand Prix, four, or 22 per cent, have been won with celebrity cases, most recently in 2008 by ‘Keep Walking’ for Johnnie Walker by BBH.

This global campaign, using a host of local stars, transformed an ailing brand into a global icon producing a total sales growth of 48 per cent and incremental sales of US$2.21bn. Cases about campaigns using a star have also won more of the other top prizes than would be expected pro rata to entries – 18 per cent of golds, and 20 per cent of silvers.

It’s interesting to contrast Ingham’s findings with data from Millward Brown, the key source of quantitative research into the use of celebrity in advertising. This is derived from its global Knowledge Point database, which contains over 100,000 test scores. As Millward Brown concludes: “The right celebrity, used in the right way, can undoubtedly be a powerful brand asset. But using a celebrity is no guarantee of effective advertising; overall, there’s very little difference between the performance of ads with celebrities versus those without.”

I believe the reason for this finding is that Millward Brown is testing what might be considered an average pool of commercials, and thus many of these will lack the outstanding strategic thinking, consumer insight, channel planning and content creation that distinguishes the elite represented by the IPA Effectiveness Awards winners. Most crucially, the average use of a celebrity will often be for relatively superficial reasons, rather than because they’re perfect casting for a role which expresses the brand idea, and amplifies and accelerates its communication.

Millward Brown also notes that there are widespread differences in the usage of celebrities between markets. The top five and bottom five countries are set out in the table below.

But just because celebrity isn’t a popular executional approach in a given territory, it doesn’t mean to say it mightn’t be very successful if applied. Indeed the relative lack of competition could even enhance the chances of success.

Similarly, Millward Brown data shows a wide divergence in the usage of stars by product category:

While there may be legal issues which militate against using celebrities to promote over-the-counter drugs, it’s hard to see any obvious reasons why they shouldn’t be deployed in the other four market sectors at the bottom of the usage table. So therein may lie an opportunity. On the other hand the top five categories are awash with stars and maybe that makes it hard to differentiate one brand campaign from another?

We have seen Dove take advantage of the quasi-generic celebrity-led approach in these categories through its campaign for real beauty using ‘ordinary’ people. In 2012 Dove won a gold in the IPA Effectiveness Awards for its shower gel campaign in China by Ogilvy & Mather Shanghai. This case showed how Dove was transformed from a small, declining player in a large market into a leading shower gel brand without relying on celebrity. The campaign grew market share from 2.1 per cent to 8.8 per cent in 16 months, delivering incremental sales of £25.5m and a ROMI of 1:1.5

This point about considering the competitive set is illustrated further by the long-term trends in celebrity usage. From a line of best fit on Millward Brown data it looks as if in the UK it has been in gentle decline, but the volatility around the average of 11 per cent is marked.

This could just be agencies reacting against what they may see at a given point in time as an over-used approach, or seeing an opportunity for celebrity because it’s temporarily out of fashion, and thus enables stand-out. So it’s wise to gauge where in the cycle the UK market appears to be when considering whether to use a star, as opposed to another executional technique.

However the drift downwards may also be to do with the relative sophistication of the UK market. Here you have to work much harder to overcome innate consumer cynicism in order to create a plausible relationship between celebrity and brand. If there’s not a really good fit, the viewer will assume the celebrity is only doing it for the money, and the impact is diluted. The star can’t just be a bolt-on.

Perhaps this explains why, as we have seen, some other countries have a much higher incidence of campaigns featuring stars: these markets may just be less developed from a marketing communications point of view. They may also have cultural traditions which favour the use of famous, aspirational role models. So overall the global trend in usage is upward, reaching nearly 20 per cent of the commercials tested by Millward Brown over the past couple of years.

Drawing conclusions from these two data sources we can see that the use of celebrity in brand campaigns is a significant type of creative execution. However there are wide variations between countries and categories which present both risks and opportunities for star-led communications.

Whereas the Millward Brown data suggests that celebrity per se isn’t a silver bullet, the celebrity cases entered into the IPA Effectiveness Awards have won a disproportionate percentage of the top prizes by demonstrating outstanding ROMI. Given there are 75 celebrity cases in EASE (the IPA’s Effectiveness Awards Search Engine), there’s plenty of guidance on how to make sure your celebrity sells.

Hamish Pringle is a strategic advisor to 23red. He tweets @HamishPringle