top of page
Work Desk

Why we exist

Personalisation and Advertising Effectiveness

What do you want first, the good news or the bad news?

The final section of this foundation of knowledge explores whether the use of personalisation in digital advertising can increase ad performance effectiveness – or not. In the first chapter exploring the discourse around privacy, we found advertisers were disappointed by the loss of third-party cookies predominantly for self-interested profit retention reasons that tie personalisation with perceived effectiveness. Conversely, consumers showed relief at Google’s changing policy, having deemed the changes an opportunity to regain control over private data and stop invasive and creepy ads. But should advertisers really shun the cookie-less changes in fear of lost ad effectiveness? In this section, I weigh up the effectiveness of existing personalisation technologies from both positive and negative standpoints, plus explore alternate approaches favoured by industry figures beyond cookie-enabled hyper-targeting. 

Bad news it is...

'The subprime data crisis'

While researching for this chapter, I became pretty surprised at the disordered and contradictory opinions from the business press surrounding effectiveness in personalised digital advertising. Evidence from one perspective infers advertisers must rely on a programmatic approach if they want to survive[1], but equally, reports suggest its funding is wasted on ad fraud and inaccurate targeting[2]. With Axiom Media pointing out 1 in 5 of MarTech solutions weren’t there in 2019, it is evident there is a significant opportunity for profitable ventures made possible by a wealth of consumer data[3]. However, during a talk titled ‘The Subprime Data Crisis’, Peter Weinberg explained the industry’s fetishisation of data is shooting us in the foot because “big data may actually be bad data, instead”[4].

Social Bubble Prototype (24).png

BIG DATA
MIGHT
ACTUALLY
BE
BAD DATA

Cut and run

Social Bubble Prototype (5).png

"technology does not need vast troves of personal data stitched together...in order to succeed" 

Social Bubble Prototype (25).png

Tim Cook
Apple CEO

Tim Cook builds upon this argument, citing a reduction in trust in institutions and the media as Apple’s reason for increased investment in anti-tracking tech, explaining “technology does not need vast troves of personal data stitched together…in order to succeed”[5]. Multiple sources within the industry are celebrating the loss of cookies because personalised digital advertising is simply not worth the additional expense. A prime example of a company that benefitted from cutting their digital ad spend was P&G, who after slashing their budget by $200m “saw no change in business outcome”[6].

Their $750m cost savings which also included halfling the number of agencies they worked with, proved Dr Augustine Fou’s point that while targeting in ad tech does work, “micro-targeting does not work nearly as well as marketers think it does”
[7]. Cookies have been producing a negative return on investment, with advertisers spending 2.5 times more for targeted display ads getting "at best 30% incremental outcomes compared to spray and pay"[8]. Whether the issue lies within the technology or the reduced trust from consumers confronted with creepy personalised ads, these results do not reflect well on ad personalisation effectiveness. 

The invaluable value exchange

As Google enforce the departure of cookies in response to growing consumer demand for online privacy, it could be argued greater perceived awareness of the invaluable value exchange with pop-up ads, opt-in menus, and hyper-targeting is to blame[9]. The online experience has become simultaneously obtrusive and intrusive, feeling more than ever like a dystopic surveillance capitalist society[10]. In creating an experience where the "user has never had control over their identities", how can advertisers expect consumers to build brand trust with a growing consciousness that their private data is being manipulated for corporate profit[11]?

Social Bubble Prototype (26).png

87%

disapproved of ads
targeted by income

Social Bubble Prototype (27).png

52%

disapproved of ads
targeted by a

Despite this Hugoneric, of Teads, believes "cookieless does not mean advertising less... it's time for non-intrusive personalisation which prioritises user experience and trust as the new normal"[12]. I agree with this statement to an extent, although it highlights personalisation's use as an umbrella term. It does not tackle the issues from whom the personalisation derives or how invasive it is. For example, YouGov’s 2,000 person survey exploring attitudes toward personalisation of online experience found 87 per cent disapproved of ads targeted by income, but only 52 per cent disapproved of the use of age[13]. While the referenced research suggests personalised advertising ineffectiveness, the generalisation of personalisation may also be harmful to measuring its strength as an advertising tool. 

Now for the good news
(kind of)

The effectiveness effect

Despite a woken cultural mentality shift heightened by media coverage of privacy breaches and trending documentaries including, Netflix’s The Social Dilemma, specialised programmatic advertising firms have proven personalisation can be effective.

Social Bubble Prototype (28).png

Clinch use creative automation tools to make dynamic video and omnichannel personalisation that claims to give clients an average sales uplift of 25 per cent and 20 per cent return on ad spend[14]. Using artificial intelligence, Clinch has developed a cycle of “first-party data enrichment” that analyses the relationship between an ad’s creative content and user engagement, ultimately creating a feedback loop of data, activation, and insights[15].

amads-logo-PPT.png

Likewise, personalisation experts, A Million Ads found "51.9% of respondents say they are more likely to purchase a product...having seen or heard a personalised ad"[16]. If consumers, as A Million Ads suggest, "actively prefer personalised ads over generic ads", the industry will need to leverage the power of personalisation according to the boundaries set by consumer trust[17]

Big money market

McKinsey_&_Company-Logo.wine.png

Share of programmatic advertising has grown as a share of the US advertising market

Screenshot 2021-12-07 at 11.56.02.png

In 2015, Richard Huntington reported for Campaign why the AdTech industry was to blame for ad blocking, which I highlighted in an earlier chapter, is rife for generation z in 2021. Perhaps, walking the personalisation line on the side of the consumer will uphold A Million Ads’ optimism for positive personalised sentiment and reduce the adblocking that is “symptomatic of the lack of consensus over digital advertising techniques and formats”[18].

James Donner, head of media strategy, would second this point, explaining the focus of Decoded Advertising “is to help brands become more proactive in delivering creative experiences where the benefit and impact of personal data are clear to consumers”
[19]. Despite this transparent and purposeful trajectory, the growth rate of programmatic over the past ten years would suggest a less-consensual personalisation formula is thriving. Programmatic, according to McKinsey, “accounted for 78.4% of US spending on display and video advertising in 2020”, which shows the dominance of the format in the digital space today[20]. This figure is up from 31.2 per cent back in 2013[21]

Inconclusive

Social Bubble Prototype (29).png

IS ADVERTISING AN ART
OR
SCIENCE?

The growth of programmatic advertising epitomises the rise of big data strategy tying back to a more philosophical argument – is advertising an art or a science? As Faris Yakob weighs in, “as rationalists we believe that marketing decisions should be based on evidence but as communicators we know that facts don’t change minds”[22]. If so, why as an industry do we focus our persuasion on the science of big data?

Referring back to McKinsey’s report, they failed to conclude whether third-party data’s effect on ad revenue was positive or not. One study estimated using third-party data in advertising increased ad revenue by 4 per cent, whereas another estimated it decreased ad revenue by as much as 50 to 60 per cent
[23]. As mentioned at the start of the chapter, the contradicting evidence supporting and disregarding the effectiveness of personalisation in digital advertising adds to the complexity of a cookie-full world. The following subsection explores how qualities of personalisation could be leveraged to improve digital advertising effectiveness beyond cookies. 

A fresh approach?

It's a balancing act

In 2014, Tucker found that internet users clicked on personalised ads almost 2X more when given control over personally identifiable information due to increased perceived control[24]. In the same study, the ads that did not use personalised text but targeted after users inputted their control options remained unchanged in effectiveness[25]. Therefore, if an internet user consensually inputs private data for personalised ads, they actively want to see the result. This point ties in with Mark Riston’s commentary on British health and beauty retailer Boots. The praise from Ritson derived from an impressive strategy that incorporates both mass and target marketing informed by their heavily subscribed loyalty card[26].

Balancing the consensually targeted approach enables consumers to receive personalised activation-focused content that resonates with previous behaviour and intent. Mass marketing used in synergy with activation personalisation, however, should not be undermined for helping to build the “top of funnel” interest that “sets up the more targeted fare for faster and more effective success”
[27]. Supported by Byron Sharp’s soft spot for mass marketing, Boots’ revamped strategy suggests a control-based system for personalisation, combined with mass marketing, can help build effectiveness. 

Boots-website-logo.png
Social Bubble Prototype (30).png

Contextual's comeback

The second approach supposedly surpassing targeted personalisation in terms of effectiveness is contextual advertising. In response to the news of a cookie-less world, the praise of contextual as a non-intrusive personalisation approach “that prioritises user experience and trust is the new normal” poses a potentially effective replacement to trackers[28]. Contextual advertising personalises ads based on the webpage context - but also extends to current affairs and seasonal holidays.

During a panel for Advertising Week Europe 2021, Morley and Beuchler explained half of UK consumers prefer brands to target them on interested search rather than demographics and location
[29]. This sentiment extended to a general preference for ads to be relevant to the content they consume[30]. From a publisher perspective, the affordances of contextual advertising to develop brand partnerships can potentially create positive sentiment and ultimately intent from the perspective of the less intrusively targeted consumer[31]

Social Bubble Prototype (4).png

1/2

of UK consumers prefer brands to target them on interested search rather than demographics and location

What do you want first, the good news or the bad news?

The findings from the third and final chapter of the foundation of knowledge highlighted a pretty bleak image of personalisation in terms of effectiveness for digital advertising. As a result of reduced trust and inaccurate tracking technologies, the returns when opting for a targeted programmatic approach are minimal, if not overwhelmingly negative. However, as Clinch and A Million Ads’ claims indicate, personalisation can be leveraged in certain contexts to increase effectiveness. Plus, the ever-increasing programmatic spend indicates the industry is determined to keep a data-full targeted approach despite policy changes, suggesting a continued optimism for ad personalisation and effectiveness.

With the motivations as to why we exist, we can now explain...

[1] https://www.thedrum.com/opinion/2021/05/27/10-benefits-programmatic-advertising-add-your-media-mix

[2] https://www.warc.com/newsandopinion/news/automation_vs_waste_the_programmatic_conundrum/40252

[3] Andreou, L., Grozalsky, J., Lanzman, D., Keens, D., (2021). The convergence of Martech and Adtech and how to combine them. Advertising Week Europe. [panel talk] virtual. 10 May 2021.

[4] https://www-warc-com.arts.idm.oclc.org/content/article/warc-exclusive/audience-targeting-why-losing-cookies-may-not-matter/133918

[5] https://www.warc.com/newsandopinion/opinion/advertisings-future-past/4126

[6] https://www-warc-com.arts.idm.oclc.org/content/article/warc-exclusive/audience-targeting-why-losing-cookies-may-not-matter/133918

[7] https://www-warc-com.arts.idm.oclc.org/content/article/warc-exclusive/audience-targeting-why-losing-cookies-may-not-matter/133918

[8] https://www-warc-com.arts.idm.oclc.org/content/article/warc-exclusive/audience-targeting-why-losing-cookies-may-not-matter/133918

[9] https://www.campaignlive.co.uk/article/why-blame-ad-blocking-crisis-lies-squarely-adtech-industry/1375249

[10] Goldfarb, A., Tucker, C., (2010). ‘Online display advertising: targeting and obtrusiveness.’ Marketing Science. 30.3, pp.389-404. Version: Author’s final manuscript

[11] Beales, M., Kloots, M., Salway, L., Guindy, S., (2021). Engaging with Consumers in a cookie-less world. Advertising Week Europe. [panel talk] virtual. 10 May 2021

[12]  https://www.exchangewire.com/blog/2021/05/25/proving-the-effectiveness-of-contextual-for-a-cookieless-world/

[13] https://www.globalwitness.org/en/blog/do-people-really-want-personalised-ads-online/

[14] https://clinch.co/#About

[15] https://clinch.co/data-enrichment/

[16] https://share.hsforms.com/1SClRTSwwQQ-L1KlYQnI4Jg2h23t 

[17] https://share.hsforms.com/1SClRTSwwQQ-L1KlYQnI4Jg2h23t

[18] https://www.campaignlive.co.uk/article/why-blame-ad-blocking-crisis-lies-squarely-adtech-industry/1375249

[19] https://www.marketingdive.com/news/whats-next-ad-industry-google-delays-third-party-cookie-death-analysis/602449/

[20] https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/the-demise-of-third-party-cookies-and-identifiers

[21] https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/the-demise-of-third-party-cookies-and-identifiers

[22] https://faris.medium.com/is-advertising-art-or-science-e0f619d54671

[23] https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/the-demise-of-third-party-cookies-and-identifiers

[24] Tucker, C., (2014). Social networks, personalised advertising, and privacy controls. Journal of Marketing Research. 51.5, pp.546-562. Version: original manuscript.

[25] Tucker, C., (2014). Social networks, personalised advertising, and privacy controls. Journal of Marketing Research. 51.5, pp.546-562. Version: original manuscript.

[26] https://www.marketingweek.com/mark-ritson-boots-brand-refresh-targeting-mass-marketing/

[27] https://www.marketingweek.com/mark-ritson-boots-brand-refresh-targeting-mass-marketing/

[28] https://www.exchangewire.com/blog/2021/05/25/proving-the-effectiveness-of-contextual-for-a-cookieless-world/

[29] Morley, N., Beuchler, T., (2021). Why a Cookieless Future isn’t Half Baked. Advertising Week Europe. [panel talk] virtual. 10 May 2021

[30] Morley, N., Beuchler, T., (2021). Why a Cookieless Future isn’t Half Baked. Advertising Week Europe. [panel talk] virtual. 10 May 2021

[31] https://www.thedrum.com/news/2021/03/25/contextual-advertising-the-solution-cookie-less-future

bottom of page