Research has always divided the ad world. There are those who value it and those who don’t. Bill Bernbach famously didn’t.
The founder of highly influential American ad agency DDB in 1949 was highly dismissive of the value of research, arguing that advertising was more of an art than a science and should therefore trust gut instinct over focus groups.
The agency went on to create such groundbreaking campaigns as VW’s Think Small and Avis’s We Try Harder – and for generally challenging established norms, ignoring rules and relying on instinct.
Its rejection of research, which became a popular narrative of the agency, only added to its reputation as a challenger brand. It also fostered a philosophy that suggested its work didn’t need to be explained. It was the expert, it knew what was right and if you didn’t like it then it wasn’t the agency for you.
Why an SSE Airtricity energy audit was a game changer for Aran Woollen Mills on its net-zero journey
Getting solid legal advice early in your company’s journey is invaluable
Water pollution has no one cause but many small steps and working together can bring great change
Empowering women in pharma: MSD Ireland’s commitment to supporting diverse leadership
Clients were intrigued and only wanted to work with them more.
But times change and the advertising equivalent of the line, ‘Trust me I’m a doctor’ doesn’t work as well today, says Dave Winterlich, chief strategy officer with dentsu, an award-winning multinational media and digital marketing communications company.
“In Bernbach’s defence, maybe research wasn’t conducted with the same level of skill and professionalism as it is today,” he explains. “We understand the human mind better now and researchers are skilled at understanding how to get to the heart of the issue.”
Of course, research can be manipulated or interpreted to prove a pre-baked hypothesis. “But when it’s done well it has huge value,” says Winterlich, speaking on podcast Inside Marketing to David Cullen, chief executive and owner of Opinions Market Research, one of the country’s leading insights and research businesses.
Research is a growing industry in Ireland, according to Cullen, who is also chair of the Marketing Society of Ireland and a board member of the Association of Irish Market Research Organisations (AIMRO). The sector is currently estimated to be worth around €70 million, an all-time high, having bounced back strongly after the pandemic.
That growth is in part due to the fact that Irish research companies are increasingly working with clients in global markets, helped by the fact that they are, according to Cullen, among the nimblest in the world. “We don’t have the luxury of specialisation to work in one sector or one type of research, so we have to be chameleons when it comes to research and really have a broad perspective,” he explains.
Cullen immediately skewers the biggest flaw in Bernbach’s position. The fact is, agency people are not fully representative of the wider public, he points out.
“Almost 80 per cent of the creative industry is based in Dublin and a huge proportion of them are under the age of 35, so are they the best judge of their own work?”, he asks.
While he agrees with Bernbach that gut is important, “that gut feeling has to be informed by evidence”, he argues.
To work, research must both be good and well interpreted. That’s where the role of the skilled moderator comes in. “In a group context, the moderator is there to moderate the session, but also to distil the essence of what people are saying,” explains Cullen.
They are not there “to try to kill the idea at the outset, but to work collaboratively to develop that idea into something that’s stronger, based on how it’s being received by the group”
One of the biggest criticisms of research is what’s known as the “say/do” problem, where people say one thing in research groups but do the opposite when nobody is watching.
“It crops up all the time in areas of sustainability,” says Winterlich. “People in research groups claim that sustainability is hugely important but yet they choose fast fashion options at the till.”
However, the skill in good research lies often in the reframing of the question, says Cullen, citing the 2016 election of Donald Trump as an example.
“I remember sitting in a room full of accountants and calling it that Trump was going to win because we had carried out some opinion polls,” he explains.
He asked three polling questions. Firstly, “who are you going to vote for?”; secondly, “who do you think will win?”; and thirdly, “who do you think your neighbour will vote for?”
“In both of the latter questions the vote went to Trump. But on the question, ‘who are you going to vote for?’ Trump was losing all day long.”
As divisive a figure as Trump is, his election campaign borrowed heavily from the world of branding and advertising, says Winterlich.
Referring to Byron Sharp, academic with the Ehrenberg-Bass Institute for Marketing Science and author of How Brands Grow, he references the role ‘distinctive assets and fame’ play in the long-term success of any brand’.
“Trump may be a villain to some and a hero to others but in a Byron Sharp world, he utilises distinctive assets to create strong ‘mental availability’, whether that is the ridiculous but instantly recognisable hair or the snappy ‘Make America Great Again’ rallying cry that was tackily emblazoned across t-shirts, flags, and baseball caps by his adoring army,” says Winterlich.
Though there is a greater appreciation for research today, the discipline itself is changing, facilitated by technology.
For a start, most research panels are now online, making research more accessible.
Unfortunately, as in other industries, where there is revenue there is opportunity for illegal gains and bots are already commonly used in countries such as the USA and China to fill surveys and reap the rewards at scale.
While AI has been used in statistical surveys for years, the current rate of technological progress means that generative AI is now pretty good at predicting the outcome of polls too.
Could AI and synthetic data, generated by algorithms, do away with the need for humans in opinion polls altogether? Not so fast, suggests Cullen.
Yes it has been proven that generative AI can “pretty much” arrive at the same conclusion as a traditional poll, but it’s the “nuance of the data underpinning” it that counts, he explains.
“If you’ve got an opinion on housing versus health and that’s informing your choice in that poll, it fails to capture any of that,” he adds.
Lack of transparency around the foundational material used to train generative AI is also an issue.
Most of all, GAI can never understand the context of the world we are living in. Says Cullen: “It can’t understand the challenges clients are facing, the competitive market they are in, or the curiosity that is required to come up with new solutions.”
To hear the Inside Marketing podcase click here