Which advertising channels are best when all else is equal?


By Samuel Scott, The Promotion Fix columnist

April 29, 2019 | 15 min read

LinkedIn’s recently unveiled ‘Reactions’ will likely lead to countless marketers tracking how many ‘likes,’ ‘celebrates,’ ‘loves,’ ‘insightfuls’ and ‘curiouses’ their posts receive – as if those designations mean anything as marketing metrics.

TV and iphone

During a talk at a CMO Network event in England last month on whether most of the internet is fake, I suggested that marketers do the following experiment.

Target a city with display ads. Survey advertising metrics such as top of mind awareness, brand recall and purchase intent among people there in the real world after the campaign. Do the same with TV, print, radio, outdoor, social media and cinema. Use a control group as well as the same creative over the same time in the same geography as much as possible. Make sure the only variable is the channel.

When all else is equal, I wondered at the time, which channels perform the best at the top of the funnel? Well, today, I have obtained research from major analyst and consultancy firms that helps to answer that question. More on that in a bit.

The McNamara Fallacy of bad metrics

During the doomed Vietnam War in the 1960s and early 1970s, the United States measured success based merely on body count – maximising the difference between the number of Vietnamese and American deaths.

Eventually, that general practice became known as the McNamara Fallacy (named for the US secretary of defence, Robert McNamara, at the time). It is basing decisions on whatever arbitrary numbers one chooses to use while ignoring all other better measurements. And that leads only to both wartime and marketing defeat.

As director Ken Burns said in his 2017 10-part documentary on the Vietnam War on US public television: "When you can't measure the things that are important, you make the things you can measure important."

Today, the online marketing world has become one giant McNamara Fallacy. Many are ignoring time-tested, quality advertising metrics in favour of whatever is easier, quicker and cheaper to measure.

Out of curiosity, I recently searched for ‘marketing metrics’ in Google and found this list of 70 from analytics platform Klipfolio at the top of the results. Almost all were short-term, direct response metrics. Those long-term advertising metrics I mentioned above were not there at all.

A post last month at Social Media Examiner also suggested that marketers use measurements including likes, retweets and engagement rates. But as LinkedIn’s Reactions will likely also show, any increase in such numbers does little for actual advertising metrics.

ad metrics

According to GlaxoSmithKline EMEA media lead Jerry Daykin in Wiemer Snijders’ Eat Your Greens – one of last year’s most important advertising books – there is barely any correlation, on average, between Facebook engagement and long-term benefits.

Further, in a recent opinion column in The Drum, Minfo founder and chief executive Roland Storti argued that the marketing industry can achieve 100% attribution in TV, VOD and OTT by developing “a platform allowing for an instant response from interested consumers. The tech combined with the ability to access info could allow marketers to create 'impulse engagement'".

But this is not – in the words of the column’s headline – “making traditional media more interactive and accountable". It is the creeping advance of short-termism and assigning of direct response metrics to all marcom. And such a focus leads only to long-term losses. Not everything in marcom is about getting someone to do something immediately.

These examples – among countless others – highlight the increasing tendency of marketers to focus on the short term and measure irrelevant things. Just like Robert McNamara.

Why we see a rise in poor marcom metrics

Today’s marketing industry loves to discuss ‘performance marketing,’ ‘results-based marketing’ and ‘growth hacking’. But the first two are simply new ways to describe direct response, and the latter phrase is testing direct response over various tools and channels. But all of this leads to people caring more about the outputs of platforms rather than the thoughts of people.

When marketers use ad tech or social media networks, for example, they often limit themselves to measuring only the short-term metrics that those tools are designed to measure. Too many focus on digital measurement but never survey people in the real world even though all marcom channels – online and offline – are just ways to reach human beings.

As Bruce Buchanan, former Jetstar chief executive and current Rokt chief executive, said at a UBS Emerging Companies forum in Australia earlier this month, “the digital economy’s ability to spin-off a wild array of easy-to-capture metrics has warped companies into a false sense of what is working for business value creation”.

The benefit of long-term metrics

Measuring long-term advertising metrics – as opposed to short-term ones – can help companies to determine key insights, according to Tina Moffett, a Forrester Research senior analyst who covers B2C marketing performance measurement.

“Measuring against long-term impact helps marketers determine if their message, content and delivery is leading to positive brand perceptions and if marketing impacts brand value,” she said.

“Measuring long-term effects can help marketers understand if efforts are driving acquisition of customers with high customer lifetime value (CLV). If marketers are focused on long term as well as short term, then they will be able to acquire loyal customers with high CLV – and hopefully act as advocates for the brand.”

According to Moffett, fashion design house Kate Spade once decided to focus on the short-term through promotions and discounts but saw the negative long-term effect of customers perceiving the brand as changing from ‘high-end’ to ‘affordable’.

Too many marketers care more about what happens immediately in dashboards and spreadsheets rather than what changes occur inside people's heads over the long term. But it is our brains that tell us what to buy.

If a brand advertises effectively to all potential buyers of a category over time, then that brand may attain the greatest top of mind awareness – meaning that the brand will be the first one that consumers imagine when they think of the category. And whenever those millions of people make a purchase in the future, they will be much more likely to buy that brand.

So, in that context, here is the research – in alphabetical order – that I obtained on which channels are most effective at doing exactly that.

The Centre for Amplified Intelligence in Australia

Karen Nelson-Field, professor of media innovation at the University of Adelaide in Australia, is now the founder and executive director of the Centre for Amplified Intelligence after working as a senior researcher for the Ehrenberg-Bass Institute.

She once gave the following presentation (PDF here) comparing the advertising effects of television, Facebook and YouTube.

TV attention

Nelson-Field told me that traditional advertising measures such as purchase intent and brand recall do not account for prior usage and will be naturally skewed depending on the characteristics of the people included in a given study. She said her company’s studies incorporate that factor.

“Getting the methodology right for any cross platform work is vital,” Nelson-Field said. “So many studies we come across get it wrong, which in turn produces really bad advice to advertisers.”

“Our advice to advertisers is this: every single reach point should be considered by its overall audience quality, its ability to be seen, its ability to attract attention and its ability to deliver sales over time. In the three years we have been considering media effectiveness, TV outperforms Facebook and YouTube in all these areas.”

Ebiquity and Gain Theory in the UK

Last year, Thinkbox commissioned a study from Ebiquity and Gain Theory to evaluate long-term advertising performance and effectiveness per channel for hundreds of brands. (Note: Thinkbox does have a pro-television agenda, but the organisation uses neutral, third-party data to support its arguments.)

A summary of the results:

ebiquity and gain theory

According to this survey, 58% of advertising’s profit return is overlooked when the long term is ignored. In other words, less than half of advertising’s profit impact happens in the short term.

“The evidence suggests that TV currently gives advertisers the best scalable return on investment,” Nick Pugh, Ebiquity’s head of effectiveness, said. “For many brands, having an accurate view of long-term impact is even more important, and our research showed that TV delivers a higher ROI not just in the short term but that this advantage multiplies when looking across longer term horizons.”

McKinsey & Company in the US

McKinsey & Company shared a report with me showing that customer loyalty is vastly overrated in most B2C categories and that marketers should focus much of their efforts on maximizing reach at the top of the funnel as a result.

mckinsey & company

“Marketing geared to growing initial consideration will exploit a more diverse and wider set of consumer segments, many with limited or perhaps even no experience with the brand,” the report stated. “The name of the game is expanding your window for growth potential.”

McKinsey & Company senior vice president Michael Betz, a former chief marketing officer in the past, has worked with B2C businesses on growth transformations to develop, execute and measure new marketing strategies in both brand and direct response contexts across online and offline channels.

In an interview, he told me that McKinsey’s research shows a link between spend at the top of the funnel and sales at the bottom. (After all, businesses tend to cut long-term advertising spend whenever recessions occur – and that hurts future results when economies grow again.)

“One company I worked with had moved almost all of their spend out of brand – traditional and digital – and into digital performance marketing – such as paid search – over the previous three years,” Betz said. “We conducted analysis that showed that there was a correlation between their reduction in true top of funnel brand spend and sales. This analysis is consistent with McKinsey’s research that shows the importance of winning at the consideration stage.”

In another example, McKinsey ran a brand campaign that focused spend on national television, radio and outdoor.

“Within three months, we saw [greater than] 20% lifts in web traffic and inquiries as well as strong improvements in conversion rates on our landing pages and website,” Betz said.

In his opinion, programmatic display can be “problematic” as well.

“We did two tests where we cut out programmatic display ads entirely and did not re-allocate to other channels,” Betz added. “We did not see any change in web traffic, inquiry volume, conversion rates or any brand awareness and recall metrics.

“We even looked the potential ‘indirect’ impact programmatic ads using advanced analytics tools and could not find any impact – such as people that were served a programmatic ad were not more likely to go to our site or click on paid search or paid social ads. Based on these results, we stopped all spend on programmatic display – though we continued with paid social and with retargeting.”

SevenOne Media in Germany

SevenOne Media wanted to compare the effectiveness of the same campaign over television and online video. So, they created an experiment in 2016 (full PDF here) with the following:

- An identical ad budget and campaign duration

- Three different but similar locations to compare the two channels with a control group

- Two different brands (an existing coffee beverage and a new yoghurt)

SevenOne Media

Across all tested advertising metrics, television performed better than online media. Most interestingly, television had an even greater effect when introducing an entirely new brand to the market.

“TV ensures a big number of exposures in and also beyond the defined target group and thereby generates higher sales effects,” Guido Modenbach, SevenOne’s managing director of market intelligence, said. “Cognitive brand effects are increased stronger with TV, which supports the hypothesis that video ad exposures are differently perceived on different platforms, and in particular, a lean-back reception – as on TV – supports the impact of the ads.”

The bigger picture

While the studies I obtained often highlighted the effectiveness of television in advertising campaigns, Ian Essling, director of survey insights at Comscore, told me the important point that “the biggest challenge trying to compare which channel is ‘better’ is that each form of media has its own strengths and weaknesses.”

“Getting a great lift from a billboard ad has to be viewed in the context of ‘how many people could see that billboard?’” he said. “Variables such as reach, frequency and recency can all affect results. A high lift with a low reach could be great, but another channel which may have a lower lift but a larger reach could be just as effective for the brand.”

“At the end of the day, the best media campaigns are those that are able to reach the right targets for the brand and use good creative design that resonates with consumers, regardless of channel.”

Fair point. No one should consider these studies the be the definitive last words on the top of the funnel. Far from it.

The above research I had found so far does not include many channels that may provide different results for various products, industries, audiences and geographies.

This column is meant to be a starting point. I contacted numerous other B2C agencies, analysts, universities, brand holding companies and industry trade groups. Some said they have research on this issue but cannot release it publicly. Some said they have not performed such studies at all. Others did not respond to inquiries.

So, I invite the industry to investigate this important topic further. BBH Labs, the UK’s CIM and IPA’s EffWorks as well as Les Binet, Peter Field and Byron Sharp – among others – may specifically be interested.

My initial question at the CMO Network focused on the top of the funnel out of personal curiosity. But marketers, of course, should not put their entire budgets into TV – or any other single channel – just because it seemingly performs best in a single context. The whole funnel must be considered.

And that will be the topic of the second part to this column in the future. While I researched this piece, I obtained so many insights that I could not fit them all here. So, stay tuned for what some of the best industry research says on optimising the entire funnel.

The Promotion Fix is an exclusive biweekly column for The Drum contributed by global keynote marketing speaker Samuel Scott, a former newspaper editor and director of marketing in the high-tech industry. Follow him on Twitter. Scott is based out of Tel Aviv, Israel.


More from Marketing

View all


Industry insights

View all
Add your own content +