Safa Amirbayat is a junior behavioural analyst at Canvas8. An economics graduate from University College London, he can be found boxing or reading a novel outside of work.

Digital advertising has changed dramatically since the introduction of the first banner ads in 1994. 

 These once-static, clickable images now travel wherever is most profitable, using our data (or psychometric profiles) as the map and transforming according to what we search, what content we consume, and how we browse. But the infrastructure such advertising has come to rely on – access to people’s personal information – is changing. On the 25th of May 2018, the EU General Data Protection Regulation (GDPR) came into effect, impacting site operators around the world, with research carried out prior to the implementation finding that 69% of British social media users had already opted out or intended to opt out of ad tracking on Facebook and Twitter. 

Given the growing awareness of how personal data feeds the digital ad industry, this reaction isn’t surprising. While the media coverage of the Cambridge Analytica scandal – in which the company allegedly used illegally-obtained Facebook data to influence elections around the world – brought concerns about how social media platforms gather users’ data to the fore, many people already felt uncomfortable with online marketing. Indeed, a survey conducted in 2016 found that 79% believed they were being tracked as a result of retargeted ads. 

 As a result, more and more web users are taking precautions to hide their digital footprints, looking on with glee as AdBlock’s ticker counts away the ads it hides from their eyes.

Why exactly does digital advertising make people feel uncomfortable? What are its downfalls? And how does GDPR present an opportunity for brands seeking to reach people online? As this year’s insights partner for 

, the UK's largest festival of behavioural science, Canvas8 took the opportunity to speak to Don Marti, an open source strategist at Mozilla and former president of the Silicon Valley Linux Users Group, to understand digital advertising’s past shortcomings and possible future.

Marti explains that people judge a product according to the medium of advertisement, not just the content. This is because the medium is a signal to the viewer of the brand’s intent to sell a good product, and while the viewer cannot know this intent for sure, it’s something they can infer from the brand’s actions. This stems from ‘asymmetric information’, whereby the seller of an item has more information about it than the potential buyer. 

 So, for many brands, the purpose of an advert is not to be unique or to inform people – it’s to send a signal of credibility.

A classic example is from the used car market. Vendors of poor quality or faulty cars have a strong incentive to conceal any defects in their stock (for example, by ‘cooking’ the numbers on the odometer), leading to vehicles being sold for more than they’re worth. In the long run, buyers realise that dishonest vendors won’t sell for the proper price, and because they have no idea of the car’s worth, they assume the worst and are will only pay the price of a bad car. Consequently, honest vendors can’t make a profit and exit the market. This leaves only dishonest sellers – a ‘market for lemons’ – and confirms buyers’ initial false suspicions. 

 A way to return good-quality vendors to a market with asymmetric information is for those vendors to convince people that they’re selling a good product with a strong signal of their honesty.

“The classic example would have to be the car commercial,” says Marti. “They’re basically short films about someone driving this car along Highway 1 in California. They contain no information about the vehicles and could've been about any car.” Advertisements of this nature function as signals – the cost commitment to advertise a product signals the brand’s belief in its superiority. “The car commercials only have to convey a single message: if it were a waste of viewers’ time to test-drive this car, then it would have been a big waste of our money to rent a helicopter and get the road closed for the obviously expensive car commercial.” 

 When a brand expends such resources on advertising, people may see the product as something worth investing in. These signals often involve a significant short-term cost for the vendor, but they stand to gain in the long-term if they stay honest about the quality of their products.

A flashy TV spot sends a stronger signal of quality than an online banner

In contrast to the example of a TV commercial for a car brand, many digital advertisements are perceived as weak signals because they require fewer resources. Early digital advertisements were blunt instruments that sometimes hit the target audience, but often reached beyond. More precise targeting has made it substantially cheaper, meaning that advertisers no longer need to pay the cost of reaching those who are unlikely to buy their products.

Modern adtech shares many features with email spam from the 1990s. In addition to being relatively low-cost, it can reach the same people on different platforms, connects to a person’s purchase history, and advertisers have the final say on which audiences to reach and when.

 As a result, web users have begun to tar digital ads with the same brush they once applied to email spam. “It starts to set off the same mental filters as email spam does. ‘Oh, I've got this piece of spam, it was clearly very cheap for the advertiser to send it and it doesn't convey useful information to me,’” says Marti. 

The perceived low effort of targeted advertising means that it carries a weak signal for viewers. “Those ads don't really convey any information about the seriousness of that brand’s intentions in the market. [People can] figure out when a message is being sent to them alone or to a wider group,” explains Marti. “It's the reason why people page through magazine ads but hang up on cold calls.” 

 The ubiquity of targeted messaging is a contributing factor to the rise of ad-blocking software, with PageFair’s 2017 Global AdBlock Report finding that 14% of US users were motivated to download such tools because they were seeing ‘too many ads on webpages’, while 29% did so to stop interruptions while browsing online. 

People are feeling overexposed to digital advertising

As well as sending a weak signal, low-cost advertising can lose effectiveness over time. “If there's an ad medium in which it's hard to get someone to believe that the ad reflects a serious investment, then that medium tends to lose attention over time,” says Marti.

 This is known as ‘peak advertising’ – the observation that digital advertising’s efficacy has declined over time. Take banner adverts as an example. In 1994, the first digital display ads had a click-through rate of 78%, a figure that has fallen to just 0.05% in 2018. 

This effect is exacerbated by a growing number of adverts vying for a fixed amount of attention. “People have vast amounts of information coming in all the time, way more than they can process,” says Marti. “So, all forms are getting less and less signal from this sort of advertising and people are paying less and less attention to it.” 

 This information overload has seen the usage of ad-blocking software surge, with installations increasing by 30% in 2016 to reach 615 million devices globally.

 As the efficacy of adverts has decreased, businesses are becoming less willing to pay a high price for them. This further weakens the signal they send because brands aren’t investing as much money to send a strong signal to buyers.

As have ads taken over the internet, people have learned to shut them out

The diluted impact of digital advertising doesn’t mean that brands should give up on the medium altogether. After all, 77% of people would prefer to filter ads than block them outright, and 83% think that not all ads are bad. 

 Additionally, a survey conducted by the CMO Council in 2017 revealed that 63% of consumers in the UK, US and Canada would respond more positively to a social media ad if it appeared in a traditional advertising medium, suggesting that the platform for the content is what’s bother people. 

There is room for brands to help people filter the messages they see. For instance, Apple introduced anti-ad-tracking features to its Safari browser in 2017. 

 According to Marti, Apple is “figuring out what do we do to take the next step in this market where tracking protection is becoming more of an expected feature.” 

 Google has also responded to people’s frustration over intrusive ads, launching an in-built ad blocker for Chrome in February 2018. 

 is shunning the fragile ad-based revenue streams of traditional journalism, making money by selling its proprietary news-scouring AI, and major publishers around the world are erecting paywalls. 

If you're able to encourage, or nudge, customers into making the right data-consent choices, then those users can shift the rules of the game for themselves

Online advertising offers new opportunities. For example, by using a chatbot to communicate with people one-to-one, Just Eat experienced a conversion rate that was 266% higher than that of its social media adverts. This may seem surprising given the 

, but 37% of Americans claim to be willing to buy through a bot. Such brand communications feel more personal and less like people are being tracked through their data. 

“GDPR is a big opportunity if you use it right,” says Marti. “We can try to rebuild the existing online advertising system on top of GDPR. [Brands can] use people’s consent to targeting to help ease their ability to tell dishonest and honest sellers apart. If you're able to encourage, or nudge, customers into making the right data-consent choices, then those users can shift the rules of the game for themselves.” 

 In other words, brands that help to intercept advertisements before they reach viewers can ensure that those that 

seen carry a stronger signal, increasing trust and engagement.

Don Marti is an open source strategist at Mozilla, working on open innovation. He is the former editor of Linux Journal and has written for LWN and other publications. Don has worked in software development, technical marketing and media, and has served as president and vice president of the Silicon Valley Linux Users Group.

Advertising

Behavioural sciences

'The first-ever banner ad on the web', The Atlantic

'GDPR: 69% users to opt out of ad tracking', Accuracast

'New data on why people hate ads: too many, too intrusive, too creepy', Vieo Design

'The market for 'lemons': quality uncertainty and the market mechanism', The Quarterly Journal of Economics

Interview with Don Marti conducted by author

'This is why we can’t have nice brands', Don Marti

'The state of the blocked web: 2017 Global Adblock Report', PageFair

'The theory of peak advertising and the future of the web', Tim Hwang and Adi Kamdar

'Average display advertising clickthrough rates', Smart Insights

'Ad blocker usage is up 30% — and a popular method publishers use to thwart it isn't working', Business Insider

'Will 2019 mark the year of personal, non-intrusive advertising', The Next Web

'Digital is the least trusted media channel among consumers, says new survey', Marketing Week

'Apple is rattling ad tech with Safari’s anti-tracking moves', Digiday

'Google Chrome’s built-in ad blocker will go live on February 15th', The Verge

'Prepare for the new paywall era', The Atlantic

Online advertising has transformed from once-static, clickable images to targeted messages that follow people as they browse. Canvas8 spoke to Don Marti, a Mozilla strategist and Nudgestock speaker, to learn how this evolution has impacted the perception and effectiveness of digital ads.

How can behavioural science improve advertising?

You’re talking to me? The science of targeted ads

How can brands destigmatise data mining?

How can behavioural economics nudge for good?