Online ads are cheap and results come in from them fast but, according to Capuchin’s Patrick Fagan, we’re measuring them all wrong.

Facebook ads suck - everyone knows it.

If you talk to people, it seems to be an open secret at this point. While the media declares Facebook ads to be an insidious tool of mass persuasion on a par with CIA mind control program MK Ultra, industry insiders have a very different view. I have lost count of the number of times a client has said something like: “No, we tried Facebook ads; it was a waste of money”.

Indeed, one survey of small business owners found that almost two thirds of respondents suspected Facebook ads don’t work. With a typical click through rate of less than one percent, it’s easy to see why business owners could be frustrated. “Trying to sell via Facebook is like walking around at a party and passing out business cards,” one of the survey respondents suggested.

However, perhaps the issue is more in digital marketers’ definitions of success than in the platform’s ability to deliver. Political parties in the UK are unlikely to have spent over a million pounds in the run-up to the election on Facebook ads if they had zero return on investment.

Digital ads are a great way to reach eyeballs: more than a third of the planet are active social media users, and more than half are internet users. In this respect, platforms like Facebook are a no-brainer, with one study estimating the cost per thousand impressions to be around $7 for Facebook, compared to upwards of $35 for television ads.

Strangely, the two types of ad seem to be held to completely different standards. No one would expect consumers, having watched the latest Coca-Cola advert on TV, to jump up off the sofa and run out in their pyjamas to buy a can from the nearest vendor.

When it comes to more traditional forms of advertising, marketers appear to understand that, like the engines on an airplane, you only notice it when it’s gone. In contrast, digital ads have concrete metrics, and marketers risk falling foul of missing the wood for the trees by focusing on the readily available performance data at the cost of less easily quantifiable metrics like awareness.

The unfortunate point here is that nurturing awareness - not inciting spontaneous behaviour - is how advertising tends to work in practice; it’s about long-term brand building, more than short-term metrics.

Academic research has been consistent in showing what makes ads effective - that is, they need to build mental availability (get remembered) by building and refreshing memory networks, and those memory network associations need to be positive and appropriate (getting remembered in the right way).

For example, regarding mental availability, a study published in the Journal of Consumer Research asked participants to view and rate some disparate pictures; afterwards, they were offered their choice of bottled water[i]. What the subjects didn’t know (and what testing showed they did not explicitly recognise) was that, for some of them, some of the pictures contained a Dasani branded bottle of water. Without this priming, 25% of participants chose Dasani; with it, this rose to 57%. This simple exposure, which consciously went unnoticed, was enough to make the brand front-of-mind and bias choice.

In other words, advertising works by winning mental ‘real estate’. However, it’s also important for that real estate to be more like ‘Selfridges on Oxford Street’ than ‘Poundland behind the car park in Hull’.

In the Spreading Activation Theory of memory, memory is conceptualised as a network of interconnecting nodes, like a spiderweb. If one sees a chair, the node representing a chair is activated in memory, and this ripples out to other nodes which have been connected via association - like ‘table’ or ‘cushion’. The goal of advertising is to link a brand to as many appropriate nodes as possible - which is why you will see smiles and families in a McDonald’s advert, but never an overweight person.

Take Coca-Cola. Its Christmas advert is so successful that many people say Christmas doesn’t officially start until they see it. But why is Christmas so germane to Coca-Cola?

Christmas has the perfect memory network for Coca-Cola to hijack. It is associated with socialising and happiness, not to mention the physical sensation of coldness. Advertising at Christmas builds appropriate purchase triggers in memory: when a customer thinks they would like a cold drink, the memory activation will ripple out to Coca-Cola, nudging a sale.

This is important because goals and contexts are crucial elements of consumer decision-making. The focus on clicks for digital ads is flawed because, most of the time, the timing is simply not right to buy when encountering a Facebook ad. Very few people are in the frame of mind to buy a Kate Spade handbag, for example, when they are browsing their social feeds sat on the toilet.

Fortunately, there are ways to test the effectiveness of digital ads beyond the blunt tool of clicks, and taking into account the psychology of how advertising really works.

Eye-tracking can be administered easily through online platforms to ensure an advert is catching attention in the first place. Consumers are cognitive misers with limited attention spans - especially considering the deluge of information they are bombarded with online. Eye-tracking helps to understand if content has that ‘thumb stopping’ component. Liberty, for example, found that adjusting the design of a direct mail envelope in response to eye-tracking results helped raise voucher redemption rates by five percent to 60%.

Facial coding can also be administered simply online, to test levels of emotional engagement with the ad once it has been noticed. Mars, for example, found measuring facial reactions helped it predict an ad’s sales lift with 75%.

Crucially, implicit testing can use reaction times to measure the effect of the ad on subconscious memory structures, evaluating the impact of the advert on brand awareness as well as ensuring that the brand has been associated with positive and appropriate memory structures. Testing of implicit attitudes can help predict future behaviour of decided and undecided voters in political elections, as several studies have shown.  

Such methods are now highly scalable, providing marketers with tools to measure digital ad effectiveness beyond clicks alone. It may come as no surprise, then, that Facebook has patented eye-tracking and facial coding technology.

Of course, at the end of the day, brands want to influence behaviour: there is no point building ‘fluffy brand feel-goodery’ for its own sake, if people don’t buy the product. This is where data helps - with attribution tools and models to track each interaction. But data won’t tell us why, which is why it should always be viewed as an able servant, to borrow from Camelot Global commercial director Richard Bateson, and never the master.

Ultimately, it may be fair to say that Facebook does suck - but only because of its flagrant disregard for privacy and its alleged suppression of free speech, not for the effectiveness of its advertising platform. Advertising works by building appropriate memories in the long-term, and, with its fantastically low cost per impression, Facebook and its contemporaries may be revolutionary.


[i] Ferraro, R., Bettman, J. R., & Chartrand, T. L. (2008). The power of strangers: The effect of incidental consumer brand encounters on brand choice. Journal of Consumer Research, 35(5), 729-741