PR polls: Robust or rubbish?
Is it true that 33 per cent of Aussie drivers have a sexy name for their car? What about 47 per cent of women more interested in househunting than man hunting? Both statistics come from polls pushed out by public relations firms but are they accurate? In a feature that first appeared in Encore, Nic Christensen investigates the precision of the PR poll.
This Valentine’s Day, love might have been all around but it was so-called “earned media” that was in the air. At least that’s according to an Encore poll of Valentine’s Day-themed public relations surveys from February 14.
The strategy of using polls to generate editorial coverage is an old PR tactic but one which appears to be growing. At the heart of the issue is the quality of the data being used.
A sample of PR surveys sent to radio, newspaper and online journalists on February 14 alone shows how the strategy works. On the most romantic day of the year it wasn’t just our partners receiving dedications of love. Research by the World Society for the Protection of Animals suggested that 30 per cent of pet owners gave their furry friends a Valentine’s Day gift. The most spoiled pets? Birds, followed by dogs and cats.
And it’s not just pets being shown the love. Research by the Coles Car Survey “revealed” that a staggering 41 per cent of Australian drivers name their car with 33 per cent finding inspiration for the name from “a celebrity, a favourite sports player, an ex-girlfriend or boyfriend or a sexy name of the opposite sex”.
And while we may seek love, it appears women prefer real estate. A Valentine’s Day study from St George Bank found 47 per cent of women are more interested in househunting than man hunting, compared to 44 per cent of men. Dr Tom Morton, director of the Australian Centre for Independent Journalism, says this is not news.
“Surveys are not a news story. They are an opinion story because most of them are about people’s opinions,” says Morton. But it takes just a quick flick through a newspaper or magazine to see that these opinions often get a run.
Morton suggests that journalists need to make a clear distinction between surveys which sample public opinion on genuine public interest issues and those which are purely promotional.
Kieran Moore, CEO of Oglivy PR, says: “Often it runs the risk of being self serving.” Case in point is one example from the Valentine’s Day survey bonanza. “Thought it could be used for some banter among the team… Realistically we wouldn’t expect any commercial messaging – just the brand mentioned in relation to the research,” wrote one PR in a pitch to a major Sydney radio station.
The producer on the receiving end of the pitch told Encore: “There are very few occasions when I have run a press release such as this that a PR has sent through to me.”
“They often come to us without any understanding of our program or what we broadcast. Then they often follow up with a phone call where it sounds like they are reading from a script.”
“When we do surveys, I talk to my clients about their objectives as a brand,” says Matthew Gain, director of PR firm Edelman. “For instance, when I used to work for Microsoft and we got a survey published which said ‘according to a study by Microsoft’, that, from my point of view, did very little.”
Gain argues that for well-established companies, the value of this type of public relations strategy can be marginal because there is already a high rate of brand awareness.
“There are a lot of brands who do this kind of survey stuff because they are trying to grow their brand awareness,” he says. “If you are trying to get a brand mentioned for a company like Coles or Microsoft, I genuinely don’t see that it adds a lot of value.”
Karalee Evans, senior director with public relations agency Text 100, agrees with Gain but says survey data can be a very successful strategy for brands trying to get established. “Often it can be a massive play for a brand in not only identifying what their consumers are seeking but also their place in the marketplace,” says Evans.
The issue of questionable data made headlines in 2011 when the ABC’s Media Watch turned its spotlight on research firm McCrindle. It uncovered figures from polls conducted by the company which were said to be representative of the Australian public despite sampling, in some instances, just 100 people.
Grant Bell, director of research firm The Acid Test, says there are hard and fast rules for generating representative surveys, the first being sampling error which gauges the accuracy of a poll using the sample size and the size of the population. The panel of people who respond to the survey must also be carefully selected in order for it to be representative. Bell says: “We work with people that specialise in this and they have a sample of representatives. For example, one company we work with has 250,000 people on their panel so if we want to talk to cat owners in Victoria they can give us a sample. Because we’re doing a lot of surveys with consumers, we know from the Census that is conducted every five years what the make up is of Australia. All we have to do is make sure that whatever sample we draw matches up with what we know. With research there are quality standards and associations.”
But while these guidelines are followed by professional data crunchers, there is no such code of practice for PR firms or companies wanting to drum up an impressive sounding number that will catch the eye of a journalist. Text 100’s Evans says: “Both sides need to get around a table and talk. What do we need to do about controlling methodology?”
Ogilvy PR’s Moore agrees. “It has to be robust. It has to be done by a reputable research company. It has to be statistically viable and it has to add something to a greater conversation.”
The Acid Test’s Bell thinks a code of practice would be a good start although he feels the reader should be more informed. “People should be more discerning when they read information and whether to take it seriously or not. A code of standards would certainly help that process.”
But Edelman’s Gain places the responsibility squarely with the journalist.
“I’ve done plenty of survey releases in my life. Some have been more newsworthy than others but ultimately I’m not the person who decides the news value,” he says. “It really is up to the journalist.”
This story first appeared in the weekly edition of Encore available for iPad and Android tablets. Visit encore.com.au for a preview of the app or click below to download.
I agree with Grants comments and also suggest referencing the source and methodology will go a long way in creating the credibility any survey requires for the data to be presented with any credibility.
User ID not verified.
I’ll throw in (i) the timing (ii) the method of interviewing (iii) the response rate. Just some top-line indicators of rigour and understanding. Or, utilise a professional.
User ID not verified.