SIGNAL IN THE NOISE: ADVERTISING IN THE AGE OF DATA

Advertising is the quintessential example of an industry known for creatively embracing what’s new and next. From emerging technologies, channels, and formats to bold, go-to-market media strategies that are guided by evolving customer expectations for personalised, seamless, and omnichannel experiences, the most successful brands, agencies, and vendors just keep moving forward. 

This topic has been discussed for a number of years now from “Talking to Ourselves”, “Lee Chow Will Only Say This Once”, CP+B’s “Woodshed”, and “the Disruptor Series” and many others. What is absolutely clear is that the agency model has shifted. We all know it’s shifted. We can feel it. Our relationship with the client has shifted. Our value proposition (and perceived value) has shifted. The culprit? For the sake of brevity – is “Data”.

The introduction of “data” into our business has shifted the perception that what, once upon a time was considered alchemy, is now quantifiable. The pendulum that swings between art and science in advertising has decidedly taken a step towards science. Why? Well, for one, it’s the natural course of human progress. 

We humans have a history of decoupling and commoditizing our once lofty constructs. You may remember years ago the arduous task (and associated costs) of building a website? Today, we have Squarespace for $16 per month. 

Moreover, with the dollars attached to advertising at large, you can bet that any number of intelligent people will attempt to commoditize any number of its functions. To this end, the advent of this “Age of Data” has put all advertising practices under scrutiny.

But the backlash today being witnessed (against the traditional ad agencies of the world) is palpable. The problem appears to be that this “Age of Data” promised far more than it has delivered.

It is the natural and inevitable course of human evolution. However, being able to quantify and benchmark every consumer transaction along the customer journey is not tantamount to success. We now, arguably, have access to every metric under the sun but the data is largely meaningless. We are still pressed daily to find the signal in the noise.

This harsh reality has manifested in plateaued CX performance, digital transformations that did not deliver the expected returns, and early efforts to capitalize on new technologies and models that took a technical, rather than operational, viability path.

The larger risk may be market-based. While taking a step back to build foundation, those firms may have missed a closing window of good economic times and deferred more aggressive strategies to an economic climate that is at best mixed and, at worst, recessionary.

At the same time AI and robotics move deeper into the organization, closer to the customer, and, more profoundly, into the very makeup and operations of the company. This presents the best mechanism to drive growth - a strategically planned ecosystem that delivers value to customers throughout their life cycle. To establish a successful ecosystem, CMO's will need to thread the needle between employee experience, customer experience, brand purpose, creative, and technology, imbuing all these crucial areas with customer obsession.

Smart CMOs will undoubtedly begin pulling back on strategies that drive short-term gains at the expense of customer affinity, including dark patterns —design patterns that manipulate customers against their own interests. Meanwhile, spend will flow back into creative as the importance of differentiated branding becomes apparent in a world of digital sameness.

At the same time, technology-driven innovation — the ability to deliver new business results through opportunities discovered by continuously experimenting with technology, both emerging and established — will soon be table stakes for leading organizations.

Today, deep learning is sorting pictures posted on Snapchat, natural language processing is providing the backbone for customer service chatbots, and machine learning is helping companies accelerate product development by handling tasks from forecasting the effect of cancer drugs to helping to edit Hollywood movies.

Just imagine an advert that dynamically changes the tone of the voiceover based on the unique preferences of the viewer. The convergence of AI with human creativity and insight will transform advertising, and we’re just beginning to see what’s possible.

Artificial Intelligence allows machines to be able to carry out tasks in a way that we would consider “smart”. And, Machine Learning is based around the idea that we should just be able to give machines access to data and let them learn for themselves. Employing both, however, despite their infinite promise, has also not yet delivered real, tangible value (at least at scale or en masse.)

Yet, we are still pressed daily to find the signal in the noise. Moreover, we are still dealing with error-laden legacy data in disparate silos and clients are ill-equipped and the speed of technological change (which means we are always catching up.)

As a result, somewhere between ‘what is infinitely possible’ and ‘what is possible today’ lies the ad agency paradox today. Selling the promise of data-driven creative and personalisation at scale to clients whose platform simply will not get them there.

This paradigm shift also extends its own vernacular – now also far more focused on return on investment and short-term results. And herein lies the problem du jour. But, in the short term, humans are still the ultimate software. 

It is as if, metaphorically, someone had just invented the paintbrush. Despite, potentially, never using one, you can still see the infinite possibilities in its premise. But you can see (in this example) that the paint brush’s promise far exceeds its current application. Ultimately, this is simply the ebb and flow of all human endeavour. 

The agency of the future will undoubtedly be consumer centric, automated, transparent, collaborative, intelligent, nimble, experiential, and focused on a sprint versus a marathon approach. They can champion creative but will undoubtedly have deep expertise in strategy, consumer insights, and measurement.

Moreover, this heightened focus on the measurement will allow agencies to not just understand campaign performance, but to also understand how a brand is moving people through a journey and how advertising is fostering that movement.

With a heightened level of insight about what people think, feel, and do (after they interact with a brand’s advertising) we are simultaneously entering an advertising landscape with more immersive experiences that engage consumers on a deeper emotional level. 

One thing we do know? The importance of data and how it’s used to make changes that put consumers first cannot be understated. Agencies that pay attention to this now are sure to set themselves up for success in the years to come.

How To Tell Stories Using Data And Statistics

SUMMARY

Details how to use data and statistics in story telling by providing six golden rules for success.

·      It can be argued that the central equation underpinning success in the modern knowledge economy is: analytics + storytelling = influence – since marketing is the best established of the “moving businesses”, the equation has particular resonance in marketing.

·      Data-driven storytelling fundamentally an act of empathy, it requires the narrator to imagine what it’s like to be in the shoes and mindset of the audience.

·      Six golden rules for data storytelling are: keep it simple, yet smart, find and use only relevant data, avoid false positives, beware the curse of knowledge, know your audience, and talk human.

·      Statistics can power stories and numbers drive narratives, but only if as a storyteller you are able to resist the seductive lure of data to tell the story for you.

NEED TO KNOW

·      The storytelling potential of data transcends marketing. Data-driven storytelling offers potential right across business and society.

·      Lightly peppering narratives with just a handful of well-chosen, killer statistics is the key to data-driven storytelling success.

·      Use data in messaging with caution. The more information you deploy in arguments when you’re looking to convince others and to change people’s minds, the more likely the audience is to resist.

·      If you put all available data into the mixer and look for relationships, you’ll find a connection and be tempted to conclude it’s important, when in fact it’s a false positive. Correlation is not causation, and connections in complex systems are very rarely caused by single factors.

·      Don’t fall into the trap of using all data available. Brand storytellers should start with the reason or purpose they’re looking for data in the first place, and then choose the right data.

·      Effective data-driven storytelling is fundamentally an act of empathy and human understanding. If you can use your research, data, and statistics to show you understand those you’re looking to influence, your story is much more likely to prove effective.

INTRODUCTION

The world of brands and brand marketing is full of data. Media data, customer data, marketing performance data. Data that reveals customer behaviour, data that attributes marketing inputs to business performance outcomes, and data that explains who’s seen which commercials, where, and how often. The sheer volume of data in marketing means that practitioners are faced with a delicious paradox: it has never been more challenging to make sense of all the data that swirls around a brand, and at the same time it has never been more possible to do so.

The storytelling potential of data transcends marketing. Data-driven storytelling offers potential right across business and society. We are all, as writer Dan Pink says, in the “moving business”; the business of persuading others to take action and respond to the stories we tell. Psychologist Daniel Kahneman has crystallised what we know about human decision-making. We make decisions fast using what Kahneman calls System 1 thinking, based on our emotional response to situations and stimuli. We then justify our decisions using slower, more cognitive thought processes, using System 2 thinking. This is where the rationality of well-presented facts, numbers, and data has a critical role to play. The most powerful and influential stories are those that appeal to both our emotions and our intellect. It can be argued, therefore, that the central equation underpinning success in the modern knowledge economy is this:

ANALYTICS + STORYTELLING = INFLUENCE.

Since marketing is the best established of the “moving businesses”, the equation has particular resonance in marketing.

WHERE TO START

It’s a common misconception that using data and statistics to build more powerful and purposeful stories is about collecting and deploying as many numbers as you can to make your point. Yet social psychology 101 tells us that the more information you deploy in arguments when you’re looking to convince others to take a course of action – and particularly when you want to change people’s minds – the more likely the audience is to resist. In recent years, this has been labelled the “Project Fear” phenomenon, after the large number of data sources and data-driven scenarios the Remain campaign deployed in the run-up to the 2016 EU Referendum in the UK.

Using data and statistics as the rational underpinning to well-balanced, emotional, and above all human stories is much more about the judicious use of a small number of well-chosen data points to support an appeal or call to action. This makes data-driven storytelling fundamentally an act of empathy. It requires the narrator to imagine what it’s like to be in the shoes – the mind; the mindset – of the audience. And to realise what it feels like to be assaulted with number after data point after statistic. Browbeating the audience into submission not only doesn’t work; it’s actively counter-productive. Lightly peppering narratives with just a handful of well-chosen, killer statistics is the key to data-driven storytelling success.

ESSENTIALS

1. KEEP IT SIMPLE – YET SMART

When you’ve thoroughly researched and understood a brand, a market, or an issue, the temptation is to share your depth of knowledge and understanding. Market researchers are particularly guilty of wanting to show all their workings out – all the data that underpins the insights they bring to their clients – but they’re not the only breed of communicators who fail to keep it simple. Clients don’t pay agencies to take them through their workings out and justify why they reached the conclusions they reached. Consumers don’t want a complete rationale of the thinking – and increasingly the data, numbers, and statistics – that underpin a campaign. Both want to know what they should do in simple, simplified terms.

Keeping it simple – yet smart – avoids the storytellers no-no of exposition or backstory. How much better as a storyteller to have the audience hanging on every twist in the story and begging for more – the next storyline, the next episode – than to be bored into submission. Building simple yet smart stories with data as the rational underpinning isn’t necessarily easy to pull off. As Mark Twain, Oscar Wilde, and Winston Churchill are all said to have said: “I would have written you a shorter letter but I didn’t have the time.”

2. FIND AND USE ONLY RELEVANT DATA

There are so many different sources of data available to brands and companies today – their own data, third-party data, and publicly-available data sets – that it’s hard to choose data that is relevant to the story you’re looking to build. Brand data can include: first-party customer journey data, social and news media content, sales data, employee attitudes and behaviours. Third-party and publicly-available data can include: analyst commentary, Government demographic data, wealth and health data, weather data, academic and peer-reviewed research, crime statistics. You name it, it’s probably been counted, analysed, and shared online.

The trick in finding and using only relevant data is to identify and deploy the corner of Big Data – what the Small Data Forum podcast calls “little big data” – that’s relevant to the hypothesis you’re trying to test, and not just what happens to be available, convenient, or easy to use. Brand storytellers should use the principles set out by writer and TED Talk favourite, Simon Sinek, and “start with why”; start with the reason or purpose they’re looking for data in the first place, and then choose the right data.

3. AVOID FALSE POSITIVES

The trouble with there being lots of potential data sets available to those looking to build brand narratives is that, when you put them together, you’re likely to find connections between different variables that are most likely meaningless. There’s every possibility that, if you’re looking to prove (rather than test) a hypothesis by putting all available data into the mixer and seeing where there might be a relationship, you’ll find something and be tempted to conclude you’ve found something important when in fact it’s a false positive. Remember the statistician’s acronym GIGO: Garbage In, Garbage Out.

As creatures, we’re hard-wired to look for relationships between variables – and the simpler the connection the better. We find it satisfying to conclude “the Gulf War was all about oil” or “Brexit was caused by unjustified fear of migrants”. Trouble is, complex connections in complex systems are very rarely caused by single factors. They’re also unlikely to be directly causally-connected. More often than not there’s a hidden third cause that you’ll have completely overlooked if you haven’t first kept it simple and second found and used only relevant data. Remember also that correlation is most definitely not causation.

4. BEWARE THE CURSE OF KNOWLEDGE

When you know a lot about a subject, it’s very hard to imagine what it’s like not to know what you know. This is called the Curse of Knowledge. In his book The Sense of Style, Harvard psychologist Steve Pinker calls out academics, government officials, lawyers, financial advisors, and many of those in Big Pharma to be among the worst afflicted by this condition.

Like using too much or irrelevant data, the Curse of Knowledge is a significant turn-off for an audience. They feel talked down to – patronised – and ignorant in the face of an expert not prepared to come down to their level. Unsurprisingly, this approach – such a common failing among those who use data and statistics in their storytelling – is also counter-productive. It’s as unengaging as showing your workings out and convinces almost no one.

5. KNOW YOUR AUDIENCE

Effective data-driven storytelling is fundamentally an act of empathy and human understanding. If you can use your research, data, and statistics to show you understand those you’re looking to influence, your story is much more likely to prove effective. The dictionary definition of insight is “a profound or deep understanding of someone or something”, and if you can use data and statistics to get under the skin and into the mindset and motivations of your audience, they’re very much more likely to be receptive to your message. The case studies on Sport England and Dove, below, show how data-driven insights can spark social and market change.

6. TALK HUMAN

It’s a curious truism of corporate behaviour that many companies and brands – including some of the most successful B2C brands, and many B2B – speak a dialect that is very unlike the way that people talk. Those mandated to talk on behalf of a company – a growing pool of individuals in our social media age in which many more voices matter – often adopt a pretentious, jargon-rich way of speaking that explains little and convinces very few. When data and statistics are at the heart of the story you’re looking to build, this problem can be made worse. Corporate and brand storytellers should resist this temptation and use numbers to support not dominate their narratives. They should use a range of emotions and talk in that rarest of corporate dialects: human.

CASE STUDIES

BRITISH HEART FOUNDATION – STAYING ALIVE

The British Heart Foundation wanted to raise awareness of how to perform hands-only CPR to restart someone’s heart if they’ve had a heart attack. Using black humour and football hard man Vinnie Jones – just featured as a gangland villain in Lock Stock & Two Smoking Barrels – the BHF managed to balance the rational with the emotional in a core campaign ad full of both information and entertainment. One of the most important pieces of information they needed to convey was the pace with which first aiders need to pump the chest of a heart attack victim, which is a couple of times every second. The planners on the BHF campaign found and used relevant data – that the Bee Gees’ disco classic Staying Alive is played at 110-120bpm – and used that as the incongruent (and memorable) soundtrack to Jones’ infomercial.

SPORT ENGLAND – THIS GIRL CAN

From early in secondary school onwards, right up until after retirement age, girls and women take part in less regular sport than boys and men; there truly is a gender exercise gap. Sport England wanted to bring about sustainable behaviour change in girls and women. Their research found that one of the main drags on women’s participation in sport was the fear of being judged by others – of sweating, make-up running, flushed cheeks, wearing unflattering Lycra, being seen out of control. 85% or women who don’t exercise say it’s for fear of being judged. And they were right – at least partially. 85% of those who see women exercising do judge them – and they do so entirely positively. They admire real women really exercising. It makes them think “I should do that!” Hence the real women that have featured in every one of the campaign’s three ads, a campaign which has led 1.6m more women to take up exercise. Sport England truly does know its audience.

DOVE – CAMPAIGN FOR REAL BEAUTY

Another brand that knows its audience and that found and used relevant data as the rational underpinning of its campaigning is Dove and its 15 year-old Campaign for Real Beauty. Back in the early 2000s, most beauty brand communication was having a negative impact on women’s self-esteem and sense of self-worth. Cross-cultural research by Dove found that in 20 countries – from Thailand to Brazil, from Russia to India – just 2% of the world’s women would use the word “beautiful” to describe themselves. This gave Dove the legitimacy to take on the beauty industry on behalf of real women, to celebrate ‘real types not stereotypes’, and to feature real women – real beauty – in their category-redefining campaign. Once the data-driven insight informed new product formulation, too, sustainable conversations with teens, tweens, and women turned into growing and sustained growth and profit for the Unilever brand.

KEY TAKEAWAYS

In his introduction his book the The Signal & The Noise, Nate Silver observes: “The numbers have no way of speaking for themselves. We speak for them. We imbue them with meaning.” It’s a sentiment repeated by Cambridge statistics professor David Spiegelhalter in the introduction to his 2019 book, The Art of Statistics.

If you keep it simple yet smart with numbers, you’re off to a good narrative start. Find and use only meaningful statistics but avoid false positives when two data points or series seem to be connected. Beware the Curse of Knowledge, a fundamentally arrogant and inhuman way to share data and statistics that you know and understand but that your audience doesn’t. Data storytelling is at its heart an act of empathy that requires skilled practitioners to know their audience – to get inside their minds – and ultimately to talk that rarest of corporate and brand dialogues: human.

Statistics can power stories and numbers drive narratives. But only if as a storyteller you are able to resist the seductive lure of data to tell the story for you or – worst of all – be presented as the story itself.