Anastasia Leng: recession proofing – how creative data can drive performance for marketers

The marketing and advertising industry has a significant blind spot when it comes to ROI: the creative. A vast 84% of marketing content is visual, yet it’s the least understood and analysed element of the marketing mix.

In the last five years, a slew of studies have repeatedly shown that creative is the largest contributor to sales uplift (i.e. Nielsen, at 47%; Kantar, which sees creative as a top driver of ad effectiveness) and the most important lever in brand profitability. All of which begs the question – why aren’t marketers spending more time focusing on the creative?

Creative data is not only a major untapped competitive advantage for CMOs, but also a valuable and unexplored data source. Every creative execution contains thousands of data points (visual, auditory, and contextual) that can now, thanks to advances in computer vision technology, be extracted, clustered, and harnessed to finally understand how the creative decisions we make impact brand and sales outcomes. So, how can we use this creative data to not only make creative decisions with confidence, but measure the impact of those decisions on the bottom line?

The building blocks to creative excellence

One of the most universal creative data applications, and the first step towards measuring creative excellence, is tracking creative quality via a Creative Quality Score (CQS). Previous studies from trillions of impressions have demonstrated that there are indeed some basic creative first principles that have repeatedly been proven to move the needle on both sales and brand growth in a statistically meaningful way.
While these learnings are platform-specific, they point to the need for a CQS, a cognitive shortcut of sorts that provides a quick health check into whether your creative is set up for success, while giving marketers control over their ability to scale content with confidence.

An increase in CQS is statistically correlated with a decrease in CPM and an increase in brand recall and return on adspend (ROAS.) Yet, the average CQS for a Fortune 500 campaign is 28%, which means 70%+ of creative work is inefficient – this can mean millions in wasted advertising spend. Will optimising for CQS alone make a creative stand out with your audience? No. Creative quality is necessary but not sufficient, but it’s an important first step towards improving content and media efficiency, especially against a backdrop of rising media costs and consumer uncertainty.

BBF Digital

Another common application of creative data is measuring brand consistency across all all content. This includes understanding how often the brand’s distinctive brand assets are used and clustering that data to assign a Consistency or Distinctiveness score to every asset. While marketers are all-too-familiar with the impact of consistent branding, they have historically lacked the ability to measure brand consistency at scale across all their ads or integrate brand consistency into their overall measurement framework. This usage of creative data will enable marketers, in real-time and across all content, to determine how brand consistency correlates with their various sales and brand KPIs.

Some applications of creative data carry positive social impact, like the ability to measure content representation and accessibility. Now more than ever, consumers want to feel seen, heard, and understood in the ads they engage with. Yet recent research reveals 76% of women going through menopause do not feel represented in advertising – and they’re not alone – a quarter of Gen Z consumers feel excluded by beauty adverts.

CreativeX studied over 3,500 image and video ads by some of the world’s best-known brands and found that male characters were 1.5 times more likely to feature in professional settings than female characters. At the same time, characters with light to medium skin tones featured twice as often in professional environments. All of this represents a huge missed opportunity as brands with more representative ads saw a 44% average stock increase over two years, 69% better business performance and 83% higher preference.

There are likely hundreds of other applications of creative data that can transform our ability to scale content production and demystify how the creative decisions we make impact business. From automating the measurement of content compliance with local regulation, to determining how sustainability-led messaging affects our brand perception, to analysing how often global content is re-used by local teams, creative data is a powerful new data source that can be harnessed and applied to shed light on critical business questions.

Why we must futureproof the marketing and advertising industries

 

As the saying goes, forewarned is forearmed. Most modern marketers are looking for ways to drive media efficiency, particularly given the continued uncertainty of the economic landscape. With a potential recession looming, marketers are continuing to struggle to prove themselves – and their budgets – to their c-suite.

Despite having just a fraction of the impact of creative quality, targeting has historically been the most popular way to optimise campaigns. We spend time and money optimising where our ad is going, who it’s being shown to, and even the time of day it’ll be featured. But even when we get the right audience to the right place at the right time, the decision of what to show them remains uninformed.

Thanks to significant advances in computer vision and machine learning, marketers can now analyse their creatives at scale and harness the creative data sitting behind millions of creative executions to map creative elements to business outcomes. Much like the marketers who recognized the potential of digital first, and moved to incorporate digital data into their understanding of consumer preferences and behaviours, those who rush to exploit creative data first will forge a powerful advantage.

Anastasia Leng is CEO of CreativeX.

You May Also Like

About Staff

Leave a Reply

Your email address will not be published. Required fields are marked *

*